Report from Gonka GitHub Discussions

Proposal #74: Gonka Labs - Monthly Report No.3

gonkalabs · Updated 9/19/2026

This report is community-posted and maintained by the Core team on GitHub, not verified by PTD moderators. Independently confirm important claims.

Proposal #74: Gonka Labs - Monthly Report No.3

Hey! Gonka Labs here.

Last updates: Report No.1 (Jul 18) and Report No.2 (Aug 18).

On June 12 the community passed Proposal #74 (full text). That vote funded six months of infra, ops, and product work. This report is month 3 of 6 - about halfway through the window we raised for.

Report #1 was launches (OpenBroker, Pulse) plus the explorer QA / inference pass and the proxy box move. Report #2 was Workspaces, PTD, holders / inference honesty, DeepSeek, rpc gRPC. This one is what landed since Aug 18.

Short version since Aug 18:

300.000.000.000 (300B) tokens processed by Proxy and OpenBroker in a month.
gonka.gg - homepage AI Tokens widget, participants show this epoch's nodes, proposals list/detail + Featured widget actually renders, wallet heatmap opens the day's txs, IBC holder channels, /snap, tx share cards
proxy.gonka.gg - GLM on the pipe, v4.1 path, persist-through-rate-limits, OAuth login, file ingest now $0, 171B+ tokens in a month
OpenBroker - DevShards v4.1, GLM live, public /status, cabinet escrow line, 129B+ tokens in a month
Pulse - chart units, sentiment faces (not Fear & Greed), news mix, catch-up without a fake 24h spike. 4.3k posts / 28.4M reach as of Sep 19
rpc.gonka.gg - wallet JSON-RPC failover (one 503 no longer blanks GG Wallet / Keplr balances), honest health checks, second phone pass on the site
PTD - Report #2 hosted on PTD itself, a proposal page lists every report (not only the latest), markdown + screenshots actually render
G-Meter - eterial.ai in the probe set, GLM-5.3 Flash as the new active model, compare uses the cheapest Gonka rate (not the average)
Gonka Chat - new product, live. Wallet-to-wallet mailbox (GG Wallet / Keplr). DMs, groups, channels, invites / replies / mentions. Message button + unread badge on gonka.gg wallet pages
x402 - new product, live. Public facilitator at x402.gonka.gg. Agents / apps pay an HTTP route in GNK or bridged USDC. No accounts, no API keys. Source on GitHub.

1. gonka.gg

Explorer side of the proposal is still "V2.0": faster, honest numbers, usable on a phone. Report #1 closed the big QA pile and stood up inference after DevShards. Report #2 added holders charts / bubble map, Confirmed/All on the inference tab, Chinese, a phone pass.

This month is smaller surfaces that people actually click, plus a few numbers that were still lying.

Homepage - AI Tokens

The four growth cards on the home page include AI Tokens. Click it: chart modal, Confirmed / All, 10 / 30 / 90 days / All, jump to the Inference page.

Same rules as /network/inference (and the same ClickHouse feed the poller writes):

Confirmed (default) - billed tokens after finish: actual in + actual out
All - Confirmed plus claimed prompt size on Gonka Labs requests that are accepted but not finished yet
We do not pour other brokers' unfinished claimed / STARTED / heartbeat probes into All. That pile is synthetic and would smear gateway shares (same reason Dahl drops synthetics before counting tokens)

Widget number is the last 24 hours. Chart is UTC days; last point is that same 24h number so hover matches the card. Y-axis is compact (7.5B, not 7,500,000,000 clipped off the left).

Participants

Expand a row on /network/participants (All / current epoch). You get this epoch's ML nodes, not last epoch's confirmation list. Hosts that added a node after confirmation were missing it in the table. Weight on the list follows the live epoch the same way.

Proposals

Featured Proposals on the homepage was not rendering. It does now.
/network/proposals list + detail restyle (chips, vote strip). Titles wrap on a phone instead of getting cut.
Prevote tab dropped from this surface.
Finished proposals stay put (no live-tally flicker after the vote ended).

Wallets / holders / calculator

Wallet activity heatmap: click a day, see that day's transactions (not only a hover count).
Holders: IBC balances by channel on hover; IBC grouping so one wallet is not split into junk rows.
Reward calculator: minus amounts display correctly; HEX sell price on the USD hint.

Other

Gateway Traffic tooltip on the inference tab rewritten for the public page (no internal notes).
gonka.gg/snap - landing for the GG Wallet MetaMask snap.
Share a tx link and the card now has a short summary (not a blank title). Example: 0634bfaf…

Links: Explorer · Inference · Participants · Proposals · Holders

2. proxy.gonka.gg

Report #1 was the dedicated-box migrate, public /status, DevShards v3. Report #2 was Workspaces, MCP in chat, DeepSeek, and the agent/tool-schema repairs. This month is the next protocol step, a fourth model, a status page you can send to a customer without lying, and a couple of settings that stop common client 400s / 429s.

Quick recap: OpenAI-compatible proxy on Gonka DevShards. Keys, balances, usage, Workspaces. Public API at https://api.proxy.gonka.gg/v1. UI at https://proxy.gonka.gg. Status at https://proxy.gonka.gg/status.

Numbers (since Aug 18 / as of Sep 18):

171B tokens in a month
Peak day ~10.2B tokens
+2k users

What shipped

Serving path moved onto DevShards v4.1. Same public URL. Keys did not have to rotate. Same job as the v3 cutover in report #1: stay on the protocol the network actually runs.

GLM-5.3-Flash came on (zai-org/GLM-5.3-Flash). Docs, /status, and the models page list it. Public /v1/models still only lists a model when the live pool can take it, so a quiet GLM hour can drop the id off that list. That is pool state, not a delist.

Public status got an honesty pass: incoming vs served, host 429s vs our own errors, in-flight vs cap, confirmation PoC when hosts leave inference mid-epoch. Incoming is a thin line so an over-capacity band does not paint the whole chart red. Kimi is off that page - it still accepts tools and then returns no tool_calls, which is a capability miss, not a Proxy outage.

We raised concurrency caps on the live gateway. DeepSeek was sitting on our own in-flight ceiling, not on empty network weight. After the bump, the concurrent-429 pile mostly went away. Caps help until you push past what hosts will take.

Persist through rate limits is a Settings toggle (off by default). Hosts still return too many concurrent requests (24/24) when every slot is full. With the toggle on, the proxy keeps that request open and retries under the hood for up to two minutes. You should almost always get a response; it may just take longer. API and Chat.

DeepSeek reasoning compatibility (off by default): official reasoning_effort / thinking: {type: enabled} get mapped onto what the hosts actually honor, so reasoning_content comes back instead of leaking into content and breaking tool harnesses.

MCP on the API, not only in Chat. Opt-in in Settings. /v1/chat/completions can list and run remote HTTPS MCP tools and return the final answer. Local npx/stdio stays in the client.

Login with proxy.gonka.gg shipped: OAuth 2.0 Authorization Code + PKCE, consent screen, scopes profile / usage / api_key. Client registration is manual. We mint a fresh revocable key for that connection; we never hand out an existing user key. Docs: https://proxy.gonka.gg/docs#login. Settings has a Connected Apps row so you can revoke it.

Also this window: public models page (context, tools, reasoning, live rates), shareable Workspace URLs.

File ingest is free now - no separate attachment line. Docs (md / txt / csv / html / pdf / docx / xlsx / odt / ods) are $0. Images are $0 inside the account caps (5 files per request, 20/min, 200/day). Settings → File ingestion has the toggle and the remaining quota. MCP in chat and web search (chat + API) stay $0. Inference is still the network rate.

Links: Product · API · Status · Models · Workspaces · Docs

3. OpenBroker

Launch was discussion #1363 and Report #1. Report #2 already had DeepSeek on both gateways, refill-to-target, deposit copy, and password reset. This month is the next protocol step, a new model, and a public status page brokers can actually send to customers.

Quick recap: OpenBroker is managed DevShards-as-a-service. Register, deposit GNK, get an obk-* key, point any OpenAI client at https://api.openbroker.gonka.gg/v1 (UI at https://openbroker.gonka.gg). GNK billing with no markup. Public stats at https://openbroker.gonka.gg/stats.

https://openbroker.gonka.gg · https://api.openbroker.gonka.gg/v1 · https://openbroker.gonka.gg/stats · https://openbroker.gonka.gg/status

Numbers (as of Sep 18):

~129B tokens and ~32M requests in a month
Peak day ~7.1B tokens (Sep 12)
Last 7 days: ~30B tokens / ~8.9M requests
Last 24h: ~2.7B tokens / ~1.0M requests
Mix in that window: DeepSeek ~95B, MiniMax ~26B, Kimi ~3.3B, GLM ~1.9B (GLM only came on late)
Brokers: 142 registered (was 70 on Aug 18) - +74 signups since report #2. 77 active ledgers, 65 still pending first deposit
Live pools: 40 MiniMax / 40 DeepSeek / 24 GLM shards on the current path

What shipped

Serving path moved onto official DevShards v4.1 (/devshard/v4.1). Same public URL. Brokers did not have to rotate keys. That is the same job as the v3 cutover in report #1: stay on the protocol the network actually runs.

GLM-5.3-Flash is live (zai-org/GLM-5.3-Flash). Docs, /v1/models, maintainer targets, and the status page all list it. Public lineup right now is MiniMax, DeepSeek, and GLM.

Public status shipped: incoming vs served, error rate by model, reason breakdown, in-flight vs cap, shard counts. No host IPs, no escrow ids. Incoming is drawn as a thin gray line so an over-capacity band does not look like the whole chart is on fire. When the network is in confirmation PoC, capacity drops on the same page - that is hosts leaving inference, not OpenBroker going dark.

We raised concurrency and token caps on the live gateway, and lifted escrow targets (DeepSeek 22 to 40, GLM 12 to 24, MiniMax stays 40). DeepSeek was sitting on our own in-flight ceiling, not on empty network weight. Caps help until you push past what hosts will take - the gateway scales itself down if you get greedy, so we stopped there.

Register: you cannot jump to the OTP step with a password under 8 characters. The API already rejected it; the form was sending the email anyway.

Cabinet got a desktop / phone pass. Billing now shows the network cost of opening new escrows we raise for traffic, plus leftover lock only on shards we cannot settle back on-chain. If an escrow settles, the remainder is not billed - that line is Epoch escrow, split by your share of successful tokens that epoch. Inference is still no markup; the cabinet just stopped hiding those network fees. Updates also go to t.me/openbroker_gg. Docs: https://openbroker.gonka.gg/docs.

Still the same promise as the launch post: run new DevShard versions and new models at production scale so relays do not find the bugs first.

Links: Product · API · Stats · Status · Docs · Register · Launch

4. Pulse

Report #1 launched it (X / Instagram / YouTube / web + sentiment on Gonka LLMs). Report #2 added Reddit, TikTok, Telegram, and the "What people are saying" digest. This month is less new sources and more "can you read the board without guessing."

Live: https://pulse.gonka.gg

Numbers (as of Sep 19):

4,348 posts indexed (was 3k+ on Aug 18)
28.4M total reach · 232K engagement (was 11.4M / 129K)
Sentiment: still 100% analyzed. Latest reading 75 (was ~65)
Last 24h (volatile): 22 posts, ~491K reach, 414 engagement
Sources same as #2: X · YouTube · Instagram · web · TikTok · Reddit, plus Telegram / discourse on the side

What shipped

The big chart numbers used to be bare digits. People had to guess if 13.2K was views or posts. Activity / reach / engagement now say posts, views, engagements. Sentiment chart says index.

General sentiment gauge got sad / happy faces on the ends so it does not look like the market Fear & Greed widget sitting next to it. Dropped the "N items analyzed" line under the gauge - it was just repeating the hero count.

In the news: one publisher's weekly digest was eating the list. We still keep a couple of those posts. The rest of the slots go to other outlets (Product Science, Coin Edition, BlockchainReporter, etc.).

Founder / investor X timelines (official accounts + BitfuryGeorge) get a deeper pull. Top Creators was overweight a single viral clip.

If collectors miss a stretch, the next run updates stored view counts. It does not dump weeks of growth onto today's 24h chart. That is how you get a fake spike.

Same public read API as before (stats / feed / news / creators / telegram / discourse). Sentiment still runs on Gonka inference via proxy.gonka.gg.

Links: Pulse

5. rpc.gonka.gg

Public RPC / LCD gateway. Still free, no key required. Report #2 already had gRPC on rpc.gonka.gg:9090, the live-host cutover, archive + indexer move, Feather drain/restore, /health that does not hang on a dead sub-service, and the first mobile pass.

This month the wallets broke. Then we fixed the path they actually use.

https://rpc.gonka.gg · https://rpc.gonka.gg/endpoints · https://rpc.gonka.gg/health

What shipped

GG Wallet, Keplr, and CosmJS POST CometBFT JSON-RPC (status, abci_query) to /chain-rpc/. That is a different pool than the GET Feather path. Mid-September one community backend started answering 503. The gateway forwarded the first 503 and never tried the next node. Balances went blank even though other backends were fine.

Pool now walks every backend on 502/503/504 before it fails. broadcast_tx 500 is still not retried - we do not want a double submit. Health checks for that pool look at the real JSON-RPC status path, not a webpage that is always 200.

Phone: second layout pass on the landing / health pages after a screenshot that was still ugly (hero buttons, "how traffic flows" stack, Powered-by cards).

As of Sep 19 public /health shows 5 backends up (2 Feather + community). Same docs, same optional /access key if you want usage tracking. gRPC still rpc.gonka.gg:9090.

Geo-aware routing from the proposal list is still open. Failover and the public health page are what we spent this window on.

Links: Gateway · Docs · Health · gRPC: rpc.gonka.gg:9090 · Access (optional)

6. PTD

Proposal text called this MTD. We shipped it as PTD. Report #1 was Pulse (the media slice). Report #2 launched the dashboard: claim a passed proposal, tasks, append-only updates, Core GitHub reports ingested for claimed and unclaimed pages. This month is the report surface, not a second product.

Live: https://ptd.gonka.gg · https://ptd.gonka.gg/reports · https://ptd.gonka.gg/reports/74

What shipped

Report #2 went up on PTD on Aug 19 (/reports/74). Same public report URL as everyone else. No login. GitHub Discussions still has the thread (#1612); PTD is the copy you can share if you only want the dashboard.

A proposal page used to show one ingested thread. It now lists every report for that id, newest first. #74 shows No.2 on top and No.1 under it. Cards on the home strip and /reports jump to the right report (#report-…).

Bodies can start with a screenshot (<img>) and still parse as markdown. First time we posted #2 that way, headings / bold / links printed as raw ** and ##. Fixed the same day.

Core ingest is still the same path as report #2. As of Sep 19 the public list also has #82, #77, #51, plus both #74 reports. We did not add a separate MTD app this window.

Links: PTD · Reports · Proposal #74 reports · GitHub source

7. G-Meter

Report #2 already had DeepSeek and OpenBroker in the probe lineup, plus the AI status notes. This month is a new broker, a new active model, and a couple of honesty / speed fixes on the public page.

Live: https://meter.gonka.gg

As of Sep 19: 11 brokers on the dashboard (eterial.ai is the new one). Active models we probe: MiniMax M2.7, DeepSeek V4 Flash, GLM-5.3 Flash. Kimi is marked deprecated - the network dropped it as a PoC model. GLM-5.2 stays optional.

What shipped

eterial.ai added to the monitored set. Same checks as everyone else: uptime, latency, price, limits.

GLM-5.3 Flash is in the active lineup (catalog, probes, compare, Limits). Limits used to hide a model until the hourly run finished; now a newly added model shows up as "not measured" instead of missing.

Compare / price ladders use the lowest Gonka broker rate, not the average. That was making Gonka look more expensive than the cheapest public quote.

Dashboard loads faster - we stopped pulling the heavy live-price payload on every home-page refresh.

proxy.gonka.gg probes now hit the API host, not the marketing site. Those checks were failing for the wrong reason.

Links: Dashboard · Compare · Get started

8. Gonka Chat

New product this window. Not on the original five-item list - same extra-ship pattern as OpenBroker and Pulse in report #1.

Gonka Chat is a mailbox for Gonka wallets. Connect GG Wallet or Keplr, then write any gonka1 address or .gnk name. No gas. Identity is the address; a name is optional. Messages and files encrypt in the browser before they leave. If the other side has not opened Chat yet, they can still read after they sign in.

Live: https://chat.gonka.gg

What shipped (late August through mid-September)

Direct chats, groups, and channels (share a handle like @squad)
Files, replies, @mentions (a person or the whole room)
Invite flow: add-to-room creates an invite in their inbox. The room does not show up until they hit Join. Open invites sit in Incoming Invites
Push notifications are less flaky
UI pass on the open-chat window, new-chat flow, and phone layout
First-run safety note; a Labs system note in the inbox is read-only so it does not look like a random DM you can argue with
Public unread count (GET /api/<address>/count) so explorer wallet pages can show Message plus a badge. Click goes to Chat.

Numbers (as of Sep 18)

First wallets signed in Aug 28
49 signed-in wallets · 76 devices
Rooms: 29 DMs, 53 groups, 42 channels
225 live messages (not counting one-time Labs notes)
5 wallets active in the last 7 days

Still early. The point this month was to get a real mailbox on the same identity people already use on gonka.gg, not a second login.

Links: Chat · App · wallet pages on gonka.gg already have the Message control

9. x402

New product this window. Not on the original five-item list - same extra-ship pattern as OpenBroker, Pulse, and Chat.

x402 is an HTTP payment standard for agents. Coinbase started it; the x402 Foundation under the Linux Foundation runs it now. Usual story on other networks: agent hits a URL, gets 402 Payment Required, pays a USDC micropayment in the same request, gets the resource. No signup, no API key, no checkout page.

Gonka did not have that rail. We shipped a public facilitator so the same flow works here, in native GNK or Ethereum-bridged Circle USDC.

Live: https://x402.gonka.gg Source: https://github.com/gonkalabs/x402-gonka

Quick wording, because this sentence gets read wrong: "agents settle on Gonka" does not mean the agent is hosted on the chain. An agent is just a program. What it means is they pay for a product that accepts GNK or USDC on Gonka.

What shipped (late August through early September)

Public facilitator: /supported, /verify, /settle, /discovery/resources, paid /demo
Docs on the same host (EN / RU / ZH): get started, sellers, CLI, protocol, HTTP API
CLI in the repo: pay (sign locally, retry with PAYMENT-SIGNATURE), keygen, serve
GNK (ngonka, 9 decimals, MsgSend) and bridged USDC (CW-20, 6 decimals, contract transfer)
x402 v2, network cosmos:gonka-mainnet, scheme exact, transfer signed-tx, flow upfront
Open source on GitHub (Sep 3). git clone + go run

How it works, short:

Client hits a paid URL.
Server answers 402 with network, asset, amount, payTo.
Client signs a Gonka transfer and retries with PAYMENT-SIGNATURE.
Seller calls POST /settle here. Tx goes on-chain. Then the resource is served, plus a PAYMENT-RESPONSE with the hash.

Settlement is upfront: money moves first. The client signs the full tx. We only broadcast. We do not hold funds.

Sellers do not talk to the chain themselves. Return 402 with your accepts, then POST /settle on x402.gonka.gg after the client retries. Same hook for a shop, another agent, or a broker that wants agents to buy from it.

As of Sep 19: /health is ok, /demo still returns HTTP 402 with GNK + USDC accepts. One live GNK pay already landed.


Paid	`/demo` · 1000 ngonka (0.000001 GNK)
Payer	`gonka150c4lsmsdr23vly466lskajkfhqh7eghmyjern`
Height	5745133 · code 0
Tx	23B76D1281F9EFCC7AFC90659480CF3219A7099959C0513720769F700E9B73B2

git clone https://github.com/gonkalabs/x402-gonka.git
cd x402-gonka
go run ./cmd/x402-gonka pay --key "$GONKA_PRIVATE_KEY" --url https://x402.gonka.gg/demo

The key stays on the signer. For USDC add --asset usdc. We did not publish a USDC settle in this window - the path is there; the proof tx above is GNK.

Still early. The point this month was to put the rail on Gonka, not to wire every Labs API behind it yet.

Links: Product / docs · Get started · Sellers · Source

Proposal checklist

Keeping the proposal list honest. Status is halfway (month 3 / 6).

Item	Status	Notes
gonka.gg V2 (engine, mobile, inference accuracy)	in progress	Inference + holders in #1/#2. This month: AI Tokens widget, current-epoch nodes, proposals UX, wallet heatmap, IBC holders, `/snap`, tx share cards. More V2 still open.
proxy.gonka.gg B2B / analytics	in progress	GLM, v4.1, status honesty, persist retries, OAuth login, file ingest $0. 171B tokens / +2k users since #2
rpc.gonka.gg geo / SLA / aggregator	in progress	Aggregation + public `/health` already live. This month: wallet POST failover + honest JSON-RPC probes. Geo routing still open.
OpenBroker	shipped / ongoing	v4.1 path, GLM, public `/status`, cabinet Epoch escrow line. 77 active / 142 registered; ~129B tokens since report #2
Pulse	shipped / ongoing	Live since #1. This month: units on charts, sentiment vs F&G, news mix, catch-up without a 24h spike. 4.3k posts / 28.4M reach as of Sep 19
PTD (proposal item: MTD)	shipped / ongoing	Pulse = media, PTD = governance. Live in #2. This month: host reports on PTD, stack all reports per proposal, markdown + screenshots render. No second MTD app.
Gonka App iOS/Android	[FILL]	[FILL]
Gonkavote expansions	[FILL]	[FILL]
GNS marketplace / scam labels	[FILL]	[FILL]
G-Router	[FILL]	[FILL]
Gonka Chat (extra)	shipped / early	New this window. Wallet mailbox at chat.gonka.gg. Invites / replies / mentions / push. 49 wallets / 225 live msgs as of Sep 18. Explorer Message + unread badge.
x402 (extra)	shipped / early	New this window. Public facilitator at x402.gonka.gg. GNK + bridged USDC. Open source Sep 3. Live GNK `/demo` settle.

Still in the second half of the six months: harden what is live, finish the checklist items that are not. Next report around mid-October.

Links

Proposal: https://gonkalabs.com/proposal · on-chain #74
Report #1: https://github.com/gonka-ai/gonka/discussions/1477
Report #2: https://github.com/gonka-ai/gonka/discussions/1612
Gonka Labs: https://gonkalabs.com
Explorer: https://gonka.gg
OpenBroker: https://openbroker.gonka.gg · updates t.me/openbroker_gg
Pulse: https://pulse.gonka.gg
Proxy: https://proxy.gonka.gg
RPC: https://rpc.gonka.gg
PTD: https://ptd.gonka.gg
G-Meter: https://meter.gonka.gg
Chat: https://chat.gonka.gg
x402: https://x402.gonka.gg · github.com/gonkalabs/x402-gonka
GitHub: https://github.com/gonkalabs

Feedback in the thread is fine - what should we push in the second half from the proposal list?

Report from Gonka GitHub Discussions

Proposal #74: Gonka Labs - Monthly Report No.2

ptd-bot ptd-bot · Updated 9/1/2026

View on GitHub ↗

This report is community-posted and maintained by the Core team on GitHub, not verified by PTD moderators. Independently confirm important claims.

Hey! Gonka Labs here. Last update was Monthly Report No.1 on Jul 18.

Report #1 was the launch month: OpenBroker and Pulse went live, the explorer got a big QA + inference pass, and proxy.gonka.gg moved onto dedicated infra under real load. This month is what we added on top. New transparency dashboard, a real agent workspace on the proxy, a third model on our own DevShards, and the ops that show up when hosts leave mid-epoch or a disk fills up.

Short version of what landed since Jul 18:

proxy.gonka.gg: Workspaces live (Chat / Docs / Slides / Sheets), MCP + file parsing, DeepSeek next to MiniMax + Kimi
PTD: new product, live (proposal transparency: claim, updates, Core reports)
gonka.gg: holders bubble map + charts, inference tab honesty, live map arcs back, full Chinese, phone pass
OpenBroker: DeepSeek on both gateways, escrow auto-top-up, broker UX + billing under real relay load
Pulse: Reddit + TikTok, Telegram community pulse, reach honesty, more press
rpc.gonka.gg: public gRPC, live gateway on a new host, archive move, Feather recovery without dropping wallets
G-Meter: stayed up through cutovers, DeepSeek + OpenBroker in the probe lineup

1. proxy.gonka.gg. Workspaces + third model + agent compat

proxy.gonka.gg is our OpenAI-compatible inference proxy on top of Gonka DevShards. API keys, balances, usage, docs, chat. Public API at https://api.proxy.gonka.gg/v1. Report #1 was the dedicated-box migrate, public /status, and DevShards v3. This month we put a real product on that pipe, and we made agents plus a third model work at the same scale.

Short version

Proxy Workspaces live (Chat / Docs / Slides / Sheets, shareable)
MCP + file parsing in chat (attachments free inside Workspaces)
DeepSeek added next to MiniMax + Kimi (32 healthy escrows each)
Agent / tool-call compat: $ref/$defs, empty tool_call.id, MiniMax content shape, leftover client knobs
Public /status now shows mid-epoch confirmation PoC, so a capacity dip is not mistaken for an outage
Fallback ladder + BYOK OpenRouter as last resort

Numbers (since Jul 18)

60B+ tokens
Registered Accounts: +3,5k since Jul 18

What shipped

Workspaces are not just a playground chat. The agent writes docs, slides, sheets, and code in one session on Gonka inference. Pick a surface at https://proxy.gonka.gg/workspaces and share the finished space.

In /chat you can connect any MCP (search, external services, your own connectors). The agent calls them in-dialog. Attach images, PDF, text, HTML, XML, CSV, XLSX, etc.

The Proxy parses files sent by user (images, pdf, xlsxs, etc) and sends the model the meaning (if user enables "File Ingestion" option in the Account settings). Attachments in Workspaces stay free.

On our own DevShards: DeepSeek as the third model, same 32-escrow target as MiniMax and Kimi. Escrow refill-to-target (settlement stays off). Empty-escrow 502s are no longer the failure mode. /status shows network confirmation PoC when hosts leave inference mid-epoch. That is a rate limit, not a Proxy outage.

Compat, because Cline / agents suddenly 400'd: Gonka hosts tightened tool-schema validation. We kept the old repairs and added request rewrites so clients still work. Inline $ref / $defs on tools and response_format.json_schema. Fill empty tool_calls[].id (and matching tool results). MiniMax content becomes [{name,type,text}] on every role. reasoning_effort: max becomes high. We drop extraBody / json_mode / output_config / fallback before upstream.

Why it matters

Proposal #74 listed proxy.gonka.gg as the B2B / analytics lane. Report #1 made the pipe carry real load. This month is the product on that pipe: a shareable workspace, agents that do not 400 on strict hosts, and a third model with enough escrow that a mid-epoch host exit looks like a rate limit, not an outage. Same gateway lessons still feed OpenBroker.

Links: Product · API · Status · Workspaces

2. PTD. Proposal Transparency Dashboard

Report #1 shipped Pulse as the first slice of the proposal's transparency item (then called MTD). PTD itself was built this window and launched in August.

Quick recap: every passed on-chain proposal is listed automatically. A team can claim a page (contact required, moderator review) and then post tasks, rich updates, and discussion. Core-team GitHub category reports are ingested automatically and show for claimed and unclaimed proposals. No login to read.

Live: https://ptd.gonka.gg

What shipped

Product built late July (email + GG Wallet auth, proposal cache from rpc.gonka.gg, claim / tasks / append-only updates / discussion). https://ptd.gonka.gg went live with TLS on Jul 31.

Early August we reworked on-chain tasks on the claim form, put the full proposal text on the initiative page (claimant can edit it), and sent claims through a moderation queue. Images work in the rich editor.

Aug 6: ingest from Core's Governance Proposal Reports (Proposal #<n>: … threads). Public /reports/[id]. No auth, no claim wall. Home shows the latest 5 reports (preview image + excerpt) plus an all-reports grid.

Aug 13: OG title / description / image so Telegram (and others) get a proper card when someone shares ptd.gonka.gg or a report link.

Why it matters

Proposal #74 asked for a transparency layer so people can see what funded work is actually doing. Pulse was the media slice. PTD is the governance one: initiative pages and the reports we (and others) post on GitHub, readable without a login.

Links: PTD · Reports · GitHub source

3. gonka.gg. Explorer + inference + holders

Report #1 already had the inference tab, Model Weight Share, GPU / reward charts, and the holders page, plus the ~240 QA smash. This month is what we added or fixed on top of that.

Short version

New on holders: bubble map, holders chart, new/active account timeseries
Inference tab: Confirmed/All, 24h per-model tokens; public DevShard aggregates API
Inference map broke; we put live arcs back and made multi-broker traffic fair
Numbers that were lying got fixed (participant cap, ghost ML nodes, model-weight formula, dust "new accounts")
Full Chinese; phone pass on homepage / header / inference / GPUs
Ongoing QA (proposal votes, reward calculator, wallets, copy, etc.)

What shipped

Public DevShard aggregates on the API so partners do not scrape the UI. New broker added in the gateway panel. Live Devshard Flow "participants" was capped at 12. Uncapped it. Readable tx type labels, copy toasts, epoch countdown tooltip in the header. MCAP / FDV on the ticker. Supply + accounts rows put back in the price tooltip. AI assistant reliability / live-data pass (assistant was already there).

Chinese localization finished across the explorer. Wallet page: account details + transaction summary.

Holders: account bubble map, new/active activity charts, holders chart. Account totals made cumulative (dust transfers still count in the total).

Honesty fixes elsewhere: escrow inference duplicates (diff nonce vs id), stale ghost ML nodes no longer inflate GPU counts, reward calculator showing 0% GPU weight coverage, proposal detail voter count (QA: proposal 95, "3 vs 12"). Inference tab: Confirmed/All on 24h gateway traffic; 24h model tokens table. Model Weight Share recounts hosted PoC, not confirmation weight (the chart was already in report #1).

Phone: homepage blocks, header, blocks/rewards widget. GPU page on phone defaults to top-4 + a latest-epoch mix instead of 12 overlapping lines. Inference tab no longer overflows the viewport on a phone.

Why it matters

Explorer side of the proposal was gonka.gg V2.0: faster engine, better UI, mobile, accurate GPU/inference metrics. Report #1 closed the painful day-to-day bugs. This is the next layer of that.

Links: Explorer · Inference · GPUs · Holders · Inference map

4. OpenBroker

We already announced the launch in discussion #1363 and covered it in Report #1: dedicated box, dual-gateway router, DevShards v3, obk-* keys, GNK ledger billing, GET /v1/usage / {id}, public stats. This month is capacity + ops + broker UX under real relay load.

Quick recap: OpenBroker is managed DevShards-as-a-service. Register, deposit GNK, get an obk-* key, point any OpenAI client at https://api.openbroker.gonka.gg/v1 (UI at https://openbroker.gonka.gg). GNK billing with no markup. Public stats at https://openbroker.gonka.gg/stats.

Numbers (since Jul 18)

50B+ tokens
Brokers: 70 registered (was 48 on Jul 18) · +23 new Brokers since Jul 18
Active broker ledgers: ~6.3k GNK on deposit

What shipped

Third model: DeepSeek-V4-Flash-0731 minted on both production gateways (live Aug 14). Docs / status / canonicalize updated.

Escrow maintainer now refills MiniMax / Kimi / DeepSeek to per-model targets on v2 + v2b independently. We tightened hard opens/run + opens/day caps and added a min operator-wallet floor.

DeepSeek clients get reasoning + reasoning_content (alias so contract clients see thinking fields).

Deposit dashboard now says the address is custodial, not personal wallet, with explicit steps. Password-reset feature added.

Ongoing: keep MiniMax / Kimi / DeepSeek escrow pools topped for dual-GW HA. proxy.gonka.gg Secondary / joingonka fallback accounts keep using OpenBroker as their overflow path. Common sense.

Why it matters

OpenBroker is still doing what the launch post promised: shake out new DevShard versions and models at production scale. DeepSeek on both gateways, refill-to-target, honest billing when final-cost is missing, plus deposit / reset UX so brokers can actually get back in when something breaks.

Links: Product · API · Stats · Launch post

5. Pulse

Report #1 launched Pulse with X / Instagram / YouTube / web, 24h charts, sentiment on Gonka LLMs, and the public read API. This month is more sources, cleaner signal, and a couple of community surfaces.

Live: https://pulse.gonka.gg

Numbers (as of Aug 18)

3k+ posts indexed
11.4M+ total reach · 129K+ engagement
Sentiment: 100% analyzed (latest reading ~65)
Sources now: X · YouTube · Instagram · website · TikTok · Reddit (+ Telegram insights separately)

What shipped

Reddit + TikTok collectors. Noise / Top Creators filters so homonyms do not pollute the feed. Off-topic X "gonka" homonym filter and UTF-16 emoji sanitization in ingest. TikTok Top Creators noise filter so junk accounts do not dominate the board.

Telegram community pulse bot + discourse "talking about" digest (EN + RU), with sentiment chips / topic bars on the community panel.

More web / press: curated sources, Medium RSS, original-article discovery. Mobile + feed QA, creator profile links, GNK price archive fallbacks.

Why it matters

Proposal #74 asked for demand activation and making Gonka easier to follow. Pulse is still that layer. Two more public sources, a Telegram surface if you do not live on the dashboard, and reach numbers that do not jump when Instagram resets a counter.

Links: Pulse · API: https://api-production-960a.up.railway.app/api (stats / feed / news / creators / telegram / discourse)

6. rpc.gonka.gg

Public RPC / LCD gateway. Report #1 mentioned it in the footer. The free / no-key pool, Hyperfusion + 6block, JSON-RPC failover, CosmWasm limits, and IP scrubbing shipped in June, before that report. This month is gRPC, a live-host cutover, and not hanging wallets when a Feather falls off tip.

Short version

gRPC now on rpc.gonka.gg:9090 (TLS)
Live gateway moved to a new host. API keys + users intact, Keplr / GG wallets kept working
Archive ClickHouse + tx index pointed at the new indexer box
Drained / restored Feather nodes when they fell off tip, then decommissioned some RPC nodes from the pool without dropping the public RPC
Public /health no longer hangs if one sub-service is dead
Full mobile UI shipped

What shipped

Mobile: full mobile layout shipped.

Some RPC nodes in the pool lagged in late July. We drained the lagging nodes from the community pool, rerouted traffic, then brought them back in when they caught up.

Cosmos gRPC is forwarded on rpc.gonka.gg:9090 (TLS at the edge, our Feather nodes behind it). Weight / drain when a Feather is in recovery after an upgrade.

Live gateway cut over to a new box. Moved to better servers. ClickHouse + tx scanner follow the archive host. One Feather decommissioned. Pool still serves as it was, just faster and more stable. Health probes are time-bounded so a dead scanner cannot freeze /health or the homepage "Powered by" strip.

Why it matters

Wallets and the explorer only look healthy if the public RPC stays on tip. This stretch was: move the box, add gRPC, do not take Keplr down while we do it. So we did.

Links: Gateway · Docs · Health · gRPC: rpc.gonka.gg:9090 · Access (optional)

7. G-Meter

This month is ops hardening + QA + DeepSeek / OpenBroker / AI status.

Live: https://meter.gonka.gg

What shipped

/compare linked in the top nav. Explained aggregate "failed tests %" vs latest-run broker list (history includes older underpayment failures).

Front QA: Limits / Providers / pricing UI; live /v1/models prices, shared price-scale domain, CIS language on compare. Mobile probe-log + providers layout.

DeepSeek-V4-Flash-0731 added to the active probe lineup (catalog / probes / brokers). Trends filter dedupe. Dropped DeepSeek probes on brokers that do not list it (Hyperfusion / Gonkarouter).

OpenBroker (api.openbroker.gonka.gg) added as OpenBroker by Gonka Labs. DeepSeek via proxy.gonka.gg writes cached network + broker AI summaries (refresh after probes). At a glance opens the broker note from a message-icon modal, with plain-language notes like "billing unavailable".

Why it matters

G-Meter is how anyone can see whether a broker is actually serving, at what price, without trusting a landing page. Adding DeepSeek and our own OpenBroker to that lineup, and not dying during a host cutover, is the point of the page.

Links: Dashboard · Compare · Get started

Still working to make Gonka Labs products better while creating new onsed. Hardening infra under real load, polishing what already shipped, building the next products on the list. More updates as they land.

Links

Feedback welcome in the thread, Gonka Labs team.

Report from Gonka GitHub Discussions

(outdated) Proposal #74: Gonka Labs - Monthly Report No. 2

nsvdev nsvdev · Updated 8/20/2026

View on GitHub ↗

This report is community-posted and maintained by the Core team on GitHub, not verified by PTD moderators. Independently confirm important claims.

Hey! Gonka Labs here. Last update was Monthly Report No.1 on Jul 18.

Report #1 was the launch month: OpenBroker and Pulse went live, the explorer got a big QA + inference pass, and proxy.gonka.gg moved onto dedicated infra under real load. This month is what we added on top. New transparency dashboard, a real agent workspace on the proxy, a third model on our own DevShards, and the ops that show up when hosts leave mid-epoch or a disk fills up.

Short version of what landed since Jul 18:

proxy.gonka.gg: Workspaces live (Chat / Docs / Slides / Sheets), MCP + file parsing, DeepSeek next to MiniMax + Kimi
PTD: new product, live (proposal transparency: claim, updates, Core reports)
gonka.gg: holders bubble map + charts, inference tab honesty, live map arcs back, full Chinese, phone pass
OpenBroker: DeepSeek on both gateways, escrow auto-top-up, broker UX + billing under real relay load
Pulse: Reddit + TikTok, Telegram community pulse, reach honesty, more press
rpc.gonka.gg: public gRPC, live gateway on a new host, archive move, Feather recovery without dropping wallets
G-Meter: stayed up through cutovers, DeepSeek + OpenBroker in the probe lineup

1. proxy.gonka.gg. Workspaces + third model + agent compat

proxy.gonka.gg is our OpenAI-compatible inference proxy on top of Gonka DevShards. API keys, balances, usage, docs, chat. Public API at https://api.proxy.gonka.gg/v1. Report #1 was the dedicated-box migrate, public /status, and DevShards v3. This month we put a real product on that pipe, and we made agents plus a third model work at the same scale.

Short version

Proxy Workspaces live (Chat / Docs / Slides / Sheets, shareable)
MCP + file parsing in chat (attachments free inside Workspaces)
DeepSeek added next to MiniMax + Kimi (32 healthy escrows each)
Agent / tool-call compat: $ref/$defs, empty tool_call.id, MiniMax content shape, leftover client knobs
Public /status now shows mid-epoch confirmation PoC, so a capacity dip is not mistaken for an outage
Fallback ladder + BYOK OpenRouter as last resort

Numbers (since Jul 18)

60B+ tokens
Registered Accounts: +3,5k since Jul 18

What shipped

Workspaces are not just a playground chat. The agent writes docs, slides, sheets, and code in one session on Gonka inference. Pick a surface at https://proxy.gonka.gg/workspaces and share the finished space.

In /chat you can connect any MCP (search, external services, your own connectors). The agent calls them in-dialog. Attach images, PDF, text, HTML, XML, CSV, XLSX, etc.

The Proxy parses files sent by user (images, pdf, xlsxs, etc) and sends the model the meaning (if user enables "File Ingestion" option in the Account settings). Attachments in Workspaces stay free.

On our own DevShards: DeepSeek as the third model, same 32-escrow target as MiniMax and Kimi. Escrow refill-to-target (settlement stays off). Empty-escrow 502s are no longer the failure mode. /status shows network confirmation PoC when hosts leave inference mid-epoch. That is a rate limit, not a Proxy outage.

Compat, because Cline / agents suddenly 400'd: Gonka hosts tightened tool-schema validation. We kept the old repairs and added request rewrites so clients still work. Inline $ref / $defs on tools and response_format.json_schema. Fill empty tool_calls[].id (and matching tool results). MiniMax content becomes [{name,type,text}] on every role. reasoning_effort: max becomes high. We drop extraBody / json_mode / output_config / fallback before upstream.

Why it matters

Proposal #74 listed proxy.gonka.gg as the B2B / analytics lane. Report #1 made the pipe carry real load. This month is the product on that pipe: a shareable workspace, agents that do not 400 on strict hosts, and a third model with enough escrow that a mid-epoch host exit looks like a rate limit, not an outage. Same gateway lessons still feed OpenBroker.

Links: Product · API · Status · Workspaces

2. PTD. Proposal Transparency Dashboard

Report #1 shipped Pulse as the first slice of the proposal's transparency item (then called MTD). PTD itself was built this window and launched in August.

Quick recap: every passed on-chain proposal is listed automatically. A team can claim a page (contact required, moderator review) and then post tasks, rich updates, and discussion. Core-team GitHub category reports are ingested automatically and show for claimed and unclaimed proposals. No login to read.

Live: https://ptd.gonka.gg

What shipped

Product built late July (email + GG Wallet auth, proposal cache from rpc.gonka.gg, claim / tasks / append-only updates / discussion). https://ptd.gonka.gg went live with TLS on Jul 31.

Early August we reworked on-chain tasks on the claim form, put the full proposal text on the initiative page (claimant can edit it), and sent claims through a moderation queue. Images work in the rich editor.

Aug 6: ingest from Core's Governance Proposal Reports (Proposal #<n>: … threads). Public /reports/[id]. No auth, no claim wall. Home shows the latest 5 reports (preview image + excerpt) plus an all-reports grid.

Aug 13: OG title / description / image so Telegram (and others) get a proper card when someone shares ptd.gonka.gg or a report link.

Why it matters

Proposal #74 asked for a transparency layer so people can see what funded work is actually doing. Pulse was the media slice. PTD is the governance one: initiative pages and the reports we (and others) post on GitHub, readable without a login.

Links: PTD · Reports · GitHub source

3. gonka.gg. Explorer + inference + holders

Report #1 already had the inference tab, Model Weight Share, GPU / reward charts, and the holders page, plus the ~240 QA smash. This month is what we added or fixed on top of that.

Short version

New on holders: bubble map, holders chart, new/active account timeseries
Inference tab: Confirmed/All, 24h per-model tokens; public DevShard aggregates API
Inference map broke; we put live arcs back and made multi-broker traffic fair
Numbers that were lying got fixed (participant cap, ghost ML nodes, model-weight formula, dust "new accounts")
Full Chinese; phone pass on homepage / header / inference / GPUs
Ongoing QA (proposal votes, reward calculator, wallets, copy, etc.)

What shipped

Public DevShard aggregates on the API so partners do not scrape the UI. New broker added in the gateway panel. Live Devshard Flow "participants" was capped at 12. Uncapped it. Readable tx type labels, copy toasts, epoch countdown tooltip in the header. MCAP / FDV on the ticker. Supply + accounts rows put back in the price tooltip. AI assistant reliability / live-data pass (assistant was already there).

Chinese localization finished across the explorer. Wallet page: account details + transaction summary.

Holders: account bubble map, new/active activity charts, holders chart. Account totals made cumulative (dust transfers still count in the total).

Honesty fixes elsewhere: escrow inference duplicates (diff nonce vs id), stale ghost ML nodes no longer inflate GPU counts, reward calculator showing 0% GPU weight coverage, proposal detail voter count (QA: proposal 95, "3 vs 12"). Inference tab: Confirmed/All on 24h gateway traffic; 24h model tokens table. Model Weight Share recounts hosted PoC, not confirmation weight (the chart was already in report #1).

Phone: homepage blocks, header, blocks/rewards widget. GPU page on phone defaults to top-4 + a latest-epoch mix instead of 12 overlapping lines. Inference tab no longer overflows the viewport on a phone.

Why it matters

Explorer side of the proposal was gonka.gg V2.0: faster engine, better UI, mobile, accurate GPU/inference metrics. Report #1 closed the painful day-to-day bugs. This is the next layer of that.

Links: Explorer · Inference · GPUs · Holders · Inference map

4. OpenBroker

We already announced the launch in discussion #1363 and covered it in Report #1: dedicated box, dual-gateway router, DevShards v3, obk-* keys, GNK ledger billing, GET /v1/usage / {id}, public stats. This month is capacity + ops + broker UX under real relay load.

Quick recap: OpenBroker is managed DevShards-as-a-service. Register, deposit GNK, get an obk-* key, point any OpenAI client at https://api.openbroker.gonka.gg/v1 (UI at https://openbroker.gonka.gg). GNK billing with no markup. Public stats at https://openbroker.gonka.gg/stats.

Numbers (since Jul 18)

50B+ tokens
Brokers: 70 registered (was 48 on Jul 18) · +23 new Brokers since Jul 18
Active broker ledgers: ~6.3k GNK on deposit

What shipped

Third model: DeepSeek-V4-Flash-0731 minted on both production gateways (live Aug 14). Docs / status / canonicalize updated.

Escrow maintainer now refills MiniMax / Kimi / DeepSeek to per-model targets on v2 + v2b independently. We tightened hard opens/run + opens/day caps and added a min operator-wallet floor.

DeepSeek clients get reasoning + reasoning_content (alias so contract clients see thinking fields).

Deposit dashboard now says the address is custodial, not personal wallet, with explicit steps. Password-reset feature added.

Ongoing: keep MiniMax / Kimi / DeepSeek escrow pools topped for dual-GW HA. proxy.gonka.gg Secondary / joingonka fallback accounts keep using OpenBroker as their overflow path. Common sense.

Why it matters

OpenBroker is still doing what the launch post promised: shake out new DevShard versions and models at production scale. DeepSeek on both gateways, refill-to-target, honest billing when final-cost is missing, plus deposit / reset UX so brokers can actually get back in when something breaks.

Links: Product · API · Stats · Launch post

5. Pulse

Report #1 launched Pulse with X / Instagram / YouTube / web, 24h charts, sentiment on Gonka LLMs, and the public read API. This month is more sources, cleaner signal, and a couple of community surfaces.

Live: https://pulse.gonka.gg

Numbers (as of Aug 18)

3k+ posts indexed
11.4M+ total reach · 129K+ engagement
Sentiment: 100% analyzed (latest reading ~65)
Sources now: X · YouTube · Instagram · website · TikTok · Reddit (+ Telegram insights separately)

What shipped

Reddit + TikTok collectors. Noise / Top Creators filters so homonyms do not pollute the feed. Off-topic X "gonka" homonym filter and UTF-16 emoji sanitization in ingest. TikTok Top Creators noise filter so junk accounts do not dominate the board.

Telegram community pulse bot + discourse "talking about" digest (EN + RU), with sentiment chips / topic bars on the community panel.

More web / press: curated sources, Medium RSS, original-article discovery. Mobile + feed QA, creator profile links, GNK price archive fallbacks.

Why it matters

Proposal #74 asked for demand activation and making Gonka easier to follow. Pulse is still that layer. Two more public sources, a Telegram surface if you do not live on the dashboard, and reach numbers that do not jump when Instagram resets a counter.

Links: Pulse · API: https://api-production-960a.up.railway.app/api (stats / feed / news / creators / telegram / discourse)

6. rpc.gonka.gg

Public RPC / LCD gateway. Report #1 mentioned it in the footer. The free / no-key pool, Hyperfusion + 6block, JSON-RPC failover, CosmWasm limits, and IP scrubbing shipped in June, before that report. This month is gRPC, a live-host cutover, and not hanging wallets when a Feather falls off tip.

Short version

gRPC now on rpc.gonka.gg:9090 (TLS)
Live gateway moved to a new host. API keys + users intact, Keplr / GG wallets kept working
Archive ClickHouse + tx index pointed at the new indexer box
Drained / restored Feather nodes when they fell off tip, then decommissioned some RPC nodes from the pool without dropping the public RPC
Public /health no longer hangs if one sub-service is dead
Full mobile UI shipped

What shipped

Mobile: full mobile layout shipped.

Some RPC nodes in the pool lagged in late July. We drained the lagging nodes from the community pool, rerouted traffic, then brought them back in when they caught up.

Cosmos gRPC is forwarded on rpc.gonka.gg:9090 (TLS at the edge, our Feather nodes behind it). Weight / drain when a Feather is in recovery after an upgrade.

Live gateway cut over to a new box. Moved to better servers. ClickHouse + tx scanner follow the archive host. One Feather decommissioned. Pool still serves as it was, just faster and more stable. Health probes are time-bounded so a dead scanner cannot freeze /health or the homepage "Powered by" strip.

Why it matters

Wallets and the explorer only look healthy if the public RPC stays on tip. This stretch was: move the box, add gRPC, do not take Keplr down while we do it. So we did.

Links: Gateway · Docs · Health · gRPC: rpc.gonka.gg:9090 · Access (optional)

7. G-Meter

This month is ops hardening + QA + DeepSeek / OpenBroker / AI status.

Live: https://meter.gonka.gg

What shipped

/compare linked in the top nav. Explained aggregate "failed tests %" vs latest-run broker list (history includes older underpayment failures).

Front QA: Limits / Providers / pricing UI; live /v1/models prices, shared price-scale domain, CIS language on compare. Mobile probe-log + providers layout.

DeepSeek-V4-Flash-0731 added to the active probe lineup (catalog / probes / brokers). Trends filter dedupe. Dropped DeepSeek probes on brokers that do not list it (Hyperfusion / Gonkarouter).

OpenBroker (api.openbroker.gonka.gg) added as OpenBroker by Gonka Labs. DeepSeek via proxy.gonka.gg writes cached network + broker AI summaries (refresh after probes). At a glance opens the broker note from a message-icon modal, with plain-language notes like "billing unavailable".

Why it matters

G-Meter is how anyone can see whether a broker is actually serving, at what price, without trusting a landing page. Adding DeepSeek and our own OpenBroker to that lineup, and not dying during a host cutover, is the point of the page.

Links: Dashboard · Compare · Get started

Still working to make Gonka Labs products better while creating new onsed. Hardening infra under real load, polishing what already shipped, building the next products on the list. More updates as they land.

Links

Feedback welcome in the thread, Gonka Labs team.

Published on PTD

Proposal #74: Gonka Labs - Monthly Report No.2

gonkalabs · Updated 8/19/2026

Posted on PTD by the proposal authors. GitHub Discussions posting was unavailable at the time.

Hey! Gonka Labs here. Last update was Monthly Report No.1 on Jul 18.

Report #1 was the launch month: OpenBroker and Pulse went live, the explorer got a big QA + inference pass, and proxy.gonka.gg moved onto dedicated infra under real load. This month is what we added on top. New transparency dashboard, a real agent workspace on the proxy, a third model on our own DevShards, and the ops that show up when hosts leave mid-epoch or a disk fills up.

Short version of what landed since Jul 18:

proxy.gonka.gg: Workspaces live (Chat / Docs / Slides / Sheets), MCP + file parsing, DeepSeek next to MiniMax + Kimi
PTD: new product, live (proposal transparency: claim, updates, Core reports)
gonka.gg: holders bubble map + charts, inference tab honesty, live map arcs back, full Chinese, phone pass
OpenBroker: DeepSeek on both gateways, escrow auto-top-up, broker UX + billing under real relay load
Pulse: Reddit + TikTok, Telegram community pulse, reach honesty, more press
rpc.gonka.gg: public gRPC, live gateway on a new host, archive move, Feather recovery without dropping wallets
G-Meter: stayed up through cutovers, DeepSeek + OpenBroker in the probe lineup

1. proxy.gonka.gg. Workspaces + third model + agent compat

proxy.gonka.gg is our OpenAI-compatible inference proxy on top of Gonka DevShards. API keys, balances, usage, docs, chat. Public API at https://api.proxy.gonka.gg/v1. Report #1 was the dedicated-box migrate, public /status, and DevShards v3. This month we put a real product on that pipe, and we made agents plus a third model work at the same scale.

Short version

Proxy Workspaces live (Chat / Docs / Slides / Sheets, shareable)
MCP + file parsing in chat (attachments free inside Workspaces)
DeepSeek added next to MiniMax + Kimi (32 healthy escrows each)
Agent / tool-call compat: $ref/$defs, empty tool_call.id, MiniMax content shape, leftover client knobs
Public /status now shows mid-epoch confirmation PoC, so a capacity dip is not mistaken for an outage
Fallback ladder + BYOK OpenRouter as last resort

Numbers (since Jul 18)

60B+ tokens
Registered Accounts: +3,5k since Jul 18

What shipped

Workspaces are not just a playground chat. The agent writes docs, slides, sheets, and code in one session on Gonka inference. Pick a surface at https://proxy.gonka.gg/workspaces and share the finished space.

In /chat you can connect any MCP (search, external services, your own connectors). The agent calls them in-dialog. Attach images, PDF, text, HTML, XML, CSV, XLSX, etc.

The Proxy parses files sent by user (images, pdf, xlsxs, etc) and sends the model the meaning (if user enables "File Ingestion" option in the Account settings). Attachments in Workspaces stay free.

On our own DevShards: DeepSeek as the third model, same 32-escrow target as MiniMax and Kimi. Escrow refill-to-target (settlement stays off). Empty-escrow 502s are no longer the failure mode. /status shows network confirmation PoC when hosts leave inference mid-epoch. That is a rate limit, not a Proxy outage.

Compat, because Cline / agents suddenly 400'd: Gonka hosts tightened tool-schema validation. We kept the old repairs and added request rewrites so clients still work. Inline $ref / $defs on tools and response_format.json_schema. Fill empty tool_calls[].id (and matching tool results). MiniMax content becomes [{name,type,text}] on every role. reasoning_effort: max becomes high. We drop extraBody / json_mode / output_config / fallback before upstream.

Why it matters

Proposal #74 listed proxy.gonka.gg as the B2B / analytics lane. Report #1 made the pipe carry real load. This month is the product on that pipe: a shareable workspace, agents that do not 400 on strict hosts, and a third model with enough escrow that a mid-epoch host exit looks like a rate limit, not an outage. Same gateway lessons still feed OpenBroker.

Links: Product · API · Status · Workspaces

2. PTD. Proposal Transparency Dashboard

Report #1 shipped Pulse as the first slice of the proposal's transparency item (then called MTD). PTD itself was built this window and launched in August.

Quick recap: every passed on-chain proposal is listed automatically. A team can claim a page (contact required, moderator review) and then post tasks, rich updates, and discussion. Core-team GitHub category reports are ingested automatically and show for claimed and unclaimed proposals. No login to read.

Live: https://ptd.gonka.gg

What shipped

Product built late July (email + GG Wallet auth, proposal cache from rpc.gonka.gg, claim / tasks / append-only updates / discussion). https://ptd.gonka.gg went live with TLS on Jul 31.

Early August we reworked on-chain tasks on the claim form, put the full proposal text on the initiative page (claimant can edit it), and sent claims through a moderation queue. Images work in the rich editor.

Aug 6: ingest from Core's Governance Proposal Reports (Proposal #<n>: … threads). Public /reports/[id]. No auth, no claim wall. Home shows the latest 5 reports (preview image + excerpt) plus an all-reports grid.

Aug 13: OG title / description / image so Telegram (and others) get a proper card when someone shares ptd.gonka.gg or a report link.

Why it matters

Proposal #74 asked for a transparency layer so people can see what funded work is actually doing. Pulse was the media slice. PTD is the governance one: initiative pages and the reports we (and others) post on GitHub, readable without a login.

Links: PTD · Reports · GitHub source

3. gonka.gg. Explorer + inference + holders

Report #1 already had the inference tab, Model Weight Share, GPU / reward charts, and the holders page, plus the ~240 QA smash. This month is what we added or fixed on top of that.

Short version

New on holders: bubble map, holders chart, new/active account timeseries
Inference tab: Confirmed/All, 24h per-model tokens; public DevShard aggregates API
Inference map broke; we put live arcs back and made multi-broker traffic fair
Numbers that were lying got fixed (participant cap, ghost ML nodes, model-weight formula, dust "new accounts")
Full Chinese; phone pass on homepage / header / inference / GPUs
Ongoing QA (proposal votes, reward calculator, wallets, copy, etc.)

What shipped

Public DevShard aggregates on the API so partners do not scrape the UI. New broker added in the gateway panel. Live Devshard Flow "participants" was capped at 12. Uncapped it. Readable tx type labels, copy toasts, epoch countdown tooltip in the header. MCAP / FDV on the ticker. Supply + accounts rows put back in the price tooltip. AI assistant reliability / live-data pass (assistant was already there).

Chinese localization finished across the explorer. Wallet page: account details + transaction summary.

Holders: account bubble map, new/active activity charts, holders chart. Account totals made cumulative (dust transfers still count in the total).

Honesty fixes elsewhere: escrow inference duplicates (diff nonce vs id), stale ghost ML nodes no longer inflate GPU counts, reward calculator showing 0% GPU weight coverage, proposal detail voter count (QA: proposal 95, "3 vs 12"). Inference tab: Confirmed/All on 24h gateway traffic; 24h model tokens table. Model Weight Share recounts hosted PoC, not confirmation weight (the chart was already in report #1).

Phone: homepage blocks, header, blocks/rewards widget. GPU page on phone defaults to top-4 + a latest-epoch mix instead of 12 overlapping lines. Inference tab no longer overflows the viewport on a phone.

Why it matters

Explorer side of the proposal was gonka.gg V2.0: faster engine, better UI, mobile, accurate GPU/inference metrics. Report #1 closed the painful day-to-day bugs. This is the next layer of that.

Links: Explorer · Inference · GPUs · Holders · Inference map

4. OpenBroker

We already announced the launch in discussion #1363 and covered it in Report #1: dedicated box, dual-gateway router, DevShards v3, obk-* keys, GNK ledger billing, GET /v1/usage / {id}, public stats. This month is capacity + ops + broker UX under real relay load.

Quick recap: OpenBroker is managed DevShards-as-a-service. Register, deposit GNK, get an obk-* key, point any OpenAI client at https://api.openbroker.gonka.gg/v1 (UI at https://openbroker.gonka.gg). GNK billing with no markup. Public stats at https://openbroker.gonka.gg/stats.

Numbers (since Jul 18)

50B+ tokens
Brokers: 70 registered (was 48 on Jul 18) · +23 new Brokers since Jul 18
Active broker ledgers: ~6.3k GNK on deposit

What shipped

Third model: DeepSeek-V4-Flash-0731 minted on both production gateways (live Aug 14). Docs / status / canonicalize updated.

Escrow maintainer now refills MiniMax / Kimi / DeepSeek to per-model targets on v2 + v2b independently. We tightened hard opens/run + opens/day caps and added a min operator-wallet floor.

DeepSeek clients get reasoning + reasoning_content (alias so contract clients see thinking fields).

Deposit dashboard now says the address is custodial, not personal wallet, with explicit steps. Password-reset feature added.

Ongoing: keep MiniMax / Kimi / DeepSeek escrow pools topped for dual-GW HA. proxy.gonka.gg Secondary / joingonka fallback accounts keep using OpenBroker as their overflow path. Common sense.

Why it matters

OpenBroker is still doing what the launch post promised: shake out new DevShard versions and models at production scale. DeepSeek on both gateways, refill-to-target, honest billing when final-cost is missing, plus deposit / reset UX so brokers can actually get back in when something breaks.

Links: Product · API · Stats · Launch post

5. Pulse

Report #1 launched Pulse with X / Instagram / YouTube / web, 24h charts, sentiment on Gonka LLMs, and the public read API. This month is more sources, cleaner signal, and a couple of community surfaces.

Live: https://pulse.gonka.gg

Numbers (as of Aug 18)

3k+ posts indexed
11.4M+ total reach · 129K+ engagement
Sentiment: 100% analyzed (latest reading ~65)
Sources now: X · YouTube · Instagram · website · TikTok · Reddit (+ Telegram insights separately)

What shipped

Reddit + TikTok collectors. Noise / Top Creators filters so homonyms do not pollute the feed. Off-topic X "gonka" homonym filter and UTF-16 emoji sanitization in ingest. TikTok Top Creators noise filter so junk accounts do not dominate the board.

Telegram community pulse bot + discourse "talking about" digest (EN + RU), with sentiment chips / topic bars on the community panel.

More web / press: curated sources, Medium RSS, original-article discovery. Mobile + feed QA, creator profile links, GNK price archive fallbacks.

Why it matters

Proposal #74 asked for demand activation and making Gonka easier to follow. Pulse is still that layer. Two more public sources, a Telegram surface if you do not live on the dashboard, and reach numbers that do not jump when Instagram resets a counter.

Links: Pulse · API: https://api-production-960a.up.railway.app/api (stats / feed / news / creators / telegram / discourse)

6. rpc.gonka.gg

Public RPC / LCD gateway. Report #1 mentioned it in the footer. The free / no-key pool, Hyperfusion + 6block, JSON-RPC failover, CosmWasm limits, and IP scrubbing shipped in June, before that report. This month is gRPC, a live-host cutover, and not hanging wallets when a Feather falls off tip.

Short version

gRPC now on rpc.gonka.gg:9090 (TLS)
Live gateway moved to a new host. API keys + users intact, Keplr / GG wallets kept working
Archive ClickHouse + tx index pointed at the new indexer box
Drained / restored Feather nodes when they fell off tip, then decommissioned some RPC nodes from the pool without dropping the public RPC
Public /health no longer hangs if one sub-service is dead
Full mobile UI shipped

What shipped

Mobile: full mobile layout shipped.

Some RPC nodes in the pool lagged in late July. We drained the lagging nodes from the community pool, rerouted traffic, then brought them back in when they caught up.

Cosmos gRPC is forwarded on rpc.gonka.gg:9090 (TLS at the edge, our Feather nodes behind it). Weight / drain when a Feather is in recovery after an upgrade.

Live gateway cut over to a new box. Moved to better servers. ClickHouse + tx scanner follow the archive host. One Feather decommissioned. Pool still serves as it was, just faster and more stable. Health probes are time-bounded so a dead scanner cannot freeze /health or the homepage "Powered by" strip.

Why it matters

Wallets and the explorer only look healthy if the public RPC stays on tip. This stretch was: move the box, add gRPC, do not take Keplr down while we do it. So we did.

Links: Gateway · Docs · Health · gRPC: rpc.gonka.gg:9090 · Access (optional)

7. G-Meter

This month is ops hardening + QA + DeepSeek / OpenBroker / AI status.

Live: https://meter.gonka.gg

What shipped

/compare linked in the top nav. Explained aggregate "failed tests %" vs latest-run broker list (history includes older underpayment failures).

Front QA: Limits / Providers / pricing UI; live /v1/models prices, shared price-scale domain, CIS language on compare. Mobile probe-log + providers layout.

DeepSeek-V4-Flash-0731 added to the active probe lineup (catalog / probes / brokers). Trends filter dedupe. Dropped DeepSeek probes on brokers that do not list it (Hyperfusion / Gonkarouter).

OpenBroker (api.openbroker.gonka.gg) added as OpenBroker by Gonka Labs. DeepSeek via proxy.gonka.gg writes cached network + broker AI summaries (refresh after probes). At a glance opens the broker note from a message-icon modal, with plain-language notes like "billing unavailable".

Why it matters

G-Meter is how anyone can see whether a broker is actually serving, at what price, without trusting a landing page. Adding DeepSeek and our own OpenBroker to that lineup, and not dying during a host cutover, is the point of the page.

Links: Dashboard · Compare · Get started

Still working to make Gonka Labs products better while creating new onsed. Hardening infra under real load, polishing what already shipped, building the next products on the list. More updates as they land.

Links

Feedback welcome in the thread, Gonka Labs team.

Report from Gonka GitHub Discussions

Proposal #74: Gonka Labs - Monthly Report No.2

gonkalabs · Updated 8/18/2026

View on GitHub ↗

This report is community-posted and maintained by the Core team on GitHub, not verified by PTD moderators. Independently confirm important claims.

Hey! Gonka Labs here. Last update was Monthly Report No.1 on Jul 18.

Report #1 was the launch month: OpenBroker and Pulse went live, the explorer got a big QA + inference pass, and proxy.gonka.gg moved onto dedicated infra under real load. This month is what we added on top. New transparency dashboard, a real agent workspace on the proxy, a third model on our own DevShards, and the ops that show up when hosts leave mid-epoch or a disk fills up.

Short version of what landed since Jul 18:

proxy.gonka.gg: Workspaces live (Chat / Docs / Slides / Sheets), MCP + file parsing, DeepSeek next to MiniMax + Kimi
PTD: new product, live (proposal transparency: claim, updates, Core reports)
gonka.gg: holders bubble map + charts, inference tab honesty, live map arcs back, full Chinese, phone pass
OpenBroker: DeepSeek on both gateways, escrow auto-top-up, broker UX + billing under real relay load
Pulse: Reddit + TikTok, Telegram community pulse, reach honesty, more press
rpc.gonka.gg: public gRPC, live gateway on a new host, archive move, Feather recovery without dropping wallets
G-Meter: stayed up through cutovers, DeepSeek + OpenBroker in the probe lineup

1. proxy.gonka.gg. Workspaces + third model + agent compat

proxy.gonka.gg is our OpenAI-compatible inference proxy on top of Gonka DevShards. API keys, balances, usage, docs, chat. Public API at https://api.proxy.gonka.gg/v1. Report #1 was the dedicated-box migrate, public /status, and DevShards v3. This month we put a real product on that pipe, and we made agents plus a third model work at the same scale.

Short version

Proxy Workspaces live (Chat / Docs / Slides / Sheets, shareable)
MCP + file parsing in chat (attachments free inside Workspaces)
DeepSeek added next to MiniMax + Kimi (32 healthy escrows each)
Agent / tool-call compat: $ref/$defs, empty tool_call.id, MiniMax content shape, leftover client knobs
Public /status now shows mid-epoch confirmation PoC, so a capacity dip is not mistaken for an outage
Fallback ladder + BYOK OpenRouter as last resort

Numbers (since Jul 18)

60B+ tokens
Registered Accounts: +3,5k since Jul 18

What shipped

Workspaces are not just a playground chat. The agent writes docs, slides, sheets, and code in one session on Gonka inference. Pick a surface at https://proxy.gonka.gg/workspaces and share the finished space.

In /chat you can connect any MCP (search, external services, your own connectors). The agent calls them in-dialog. Attach images, PDF, text, HTML, XML, CSV, XLSX, etc.

The Proxy parses files sent by user (images, pdf, xlsxs, etc) and sends the model the meaning (if user enables "File Ingestion" option in the Account settings). Attachments in Workspaces stay free.

On our own DevShards: DeepSeek as the third model, same 32-escrow target as MiniMax and Kimi. Escrow refill-to-target (settlement stays off). Empty-escrow 502s are no longer the failure mode. /status shows network confirmation PoC when hosts leave inference mid-epoch. That is a rate limit, not a Proxy outage.

Compat, because Cline / agents suddenly 400'd: Gonka hosts tightened tool-schema validation. We kept the old repairs and added request rewrites so clients still work. Inline $ref / $defs on tools and response_format.json_schema. Fill empty tool_calls[].id (and matching tool results). MiniMax content becomes [{name,type,text}] on every role. reasoning_effort: max becomes high. We drop extraBody / json_mode / output_config / fallback before upstream.

Why it matters

Proposal #74 listed proxy.gonka.gg as the B2B / analytics lane. Report #1 made the pipe carry real load. This month is the product on that pipe: a shareable workspace, agents that do not 400 on strict hosts, and a third model with enough escrow that a mid-epoch host exit looks like a rate limit, not an outage. Same gateway lessons still feed OpenBroker.

Links: Product · API · Status · Workspaces

2. PTD. Proposal Transparency Dashboard

Report #1 shipped Pulse as the first slice of the proposal's transparency item (then called MTD). PTD itself was built this window and launched in August.

Quick recap: every passed on-chain proposal is listed automatically. A team can claim a page (contact required, moderator review) and then post tasks, rich updates, and discussion. Core-team GitHub category reports are ingested automatically and show for claimed and unclaimed proposals. No login to read.

Live: https://ptd.gonka.gg

What shipped

Product built late July (email + GG Wallet auth, proposal cache from rpc.gonka.gg, claim / tasks / append-only updates / discussion). https://ptd.gonka.gg went live with TLS on Jul 31.

Early August we reworked on-chain tasks on the claim form, put the full proposal text on the initiative page (claimant can edit it), and sent claims through a moderation queue. Images work in the rich editor.

Aug 6: ingest from Core's Governance Proposal Reports (Proposal #<n>: … threads). Public /reports/[id]. No auth, no claim wall. Home shows the latest 5 reports (preview image + excerpt) plus an all-reports grid.

Aug 13: OG title / description / image so Telegram (and others) get a proper card when someone shares ptd.gonka.gg or a report link.

Why it matters

Proposal #74 asked for a transparency layer so people can see what funded work is actually doing. Pulse was the media slice. PTD is the governance one: initiative pages and the reports we (and others) post on GitHub, readable without a login.

Links: PTD · Reports · GitHub source

3. gonka.gg. Explorer + inference + holders

Report #1 already had the inference tab, Model Weight Share, GPU / reward charts, and the holders page, plus the ~240 QA smash. This month is what we added or fixed on top of that.

Short version

New on holders: bubble map, holders chart, new/active account timeseries
Inference tab: Confirmed/All, 24h per-model tokens; public DevShard aggregates API
Inference map broke; we put live arcs back and made multi-broker traffic fair
Numbers that were lying got fixed (participant cap, ghost ML nodes, model-weight formula, dust "new accounts")
Full Chinese; phone pass on homepage / header / inference / GPUs
Ongoing QA (proposal votes, reward calculator, wallets, copy, etc.)

What shipped

Public DevShard aggregates on the API so partners do not scrape the UI. New broker added in the gateway panel. Live Devshard Flow "participants" was capped at 12. Uncapped it. Readable tx type labels, copy toasts, epoch countdown tooltip in the header. MCAP / FDV on the ticker. Supply + accounts rows put back in the price tooltip. AI assistant reliability / live-data pass (assistant was already there).

Chinese localization finished across the explorer. Wallet page: account details + transaction summary.

Holders: account bubble map, new/active activity charts, holders chart. Account totals made cumulative (dust transfers still count in the total).

Honesty fixes elsewhere: escrow inference duplicates (diff nonce vs id), stale ghost ML nodes no longer inflate GPU counts, reward calculator showing 0% GPU weight coverage, proposal detail voter count (QA: proposal 95, "3 vs 12"). Inference tab: Confirmed/All on 24h gateway traffic; 24h model tokens table. Model Weight Share recounts hosted PoC, not confirmation weight (the chart was already in report #1).

Phone: homepage blocks, header, blocks/rewards widget. GPU page on phone defaults to top-4 + a latest-epoch mix instead of 12 overlapping lines. Inference tab no longer overflows the viewport on a phone.

Why it matters

Explorer side of the proposal was gonka.gg V2.0: faster engine, better UI, mobile, accurate GPU/inference metrics. Report #1 closed the painful day-to-day bugs. This is the next layer of that.

Links: Explorer · Inference · GPUs · Holders · Inference map

4. OpenBroker

We already announced the launch in discussion #1363 and covered it in Report #1: dedicated box, dual-gateway router, DevShards v3, obk-* keys, GNK ledger billing, GET /v1/usage / {id}, public stats. This month is capacity + ops + broker UX under real relay load.

Quick recap: OpenBroker is managed DevShards-as-a-service. Register, deposit GNK, get an obk-* key, point any OpenAI client at https://api.openbroker.gonka.gg/v1 (UI at https://openbroker.gonka.gg). GNK billing with no markup. Public stats at https://openbroker.gonka.gg/stats.

Numbers (since Jul 18)

50B+ tokens
Brokers: 70 registered (was 48 on Jul 18) · +23 new Brokers since Jul 18
Active broker ledgers: ~6.3k GNK on deposit

What shipped

Third model: DeepSeek-V4-Flash-0731 minted on both production gateways (live Aug 14). Docs / status / canonicalize updated.

Escrow maintainer now refills MiniMax / Kimi / DeepSeek to per-model targets on v2 + v2b independently. We tightened hard opens/run + opens/day caps and added a min operator-wallet floor.

DeepSeek clients get reasoning + reasoning_content (alias so contract clients see thinking fields).

Deposit dashboard now says the address is custodial, not personal wallet, with explicit steps. Password-reset feature added.

Ongoing: keep MiniMax / Kimi / DeepSeek escrow pools topped for dual-GW HA. proxy.gonka.gg Secondary / joingonka fallback accounts keep using OpenBroker as their overflow path. Common sense.

Why it matters

OpenBroker is still doing what the launch post promised: shake out new DevShard versions and models at production scale. DeepSeek on both gateways, refill-to-target, honest billing when final-cost is missing, plus deposit / reset UX so brokers can actually get back in when something breaks.

Links: Product · API · Stats · Launch post

5. Pulse

Report #1 launched Pulse with X / Instagram / YouTube / web, 24h charts, sentiment on Gonka LLMs, and the public read API. This month is more sources, cleaner signal, and a couple of community surfaces.

Live: https://pulse.gonka.gg

Numbers (as of Aug 18)

3k+ posts indexed
11.4M+ total reach · 129K+ engagement
Sentiment: 100% analyzed (latest reading ~65)
Sources now: X · YouTube · Instagram · website · TikTok · Reddit (+ Telegram insights separately)

What shipped

Reddit + TikTok collectors. Noise / Top Creators filters so homonyms do not pollute the feed. Off-topic X "gonka" homonym filter and UTF-16 emoji sanitization in ingest. TikTok Top Creators noise filter so junk accounts do not dominate the board.

Telegram community pulse bot + discourse "talking about" digest (EN + RU), with sentiment chips / topic bars on the community panel.

More web / press: curated sources, Medium RSS, original-article discovery. Mobile + feed QA, creator profile links, GNK price archive fallbacks.

Why it matters

Proposal #74 asked for demand activation and making Gonka easier to follow. Pulse is still that layer. Two more public sources, a Telegram surface if you do not live on the dashboard, and reach numbers that do not jump when Instagram resets a counter.

Links: Pulse · API: https://api-production-960a.up.railway.app/api (stats / feed / news / creators / telegram / discourse)

6. rpc.gonka.gg

Public RPC / LCD gateway. Report #1 mentioned it in the footer. The free / no-key pool, Hyperfusion + 6block, JSON-RPC failover, CosmWasm limits, and IP scrubbing shipped in June, before that report. This month is gRPC, a live-host cutover, and not hanging wallets when a Feather falls off tip.

Short version

gRPC now on rpc.gonka.gg:9090 (TLS)
Live gateway moved to a new host. API keys + users intact, Keplr / GG wallets kept working
Archive ClickHouse + tx index pointed at the new indexer box
Drained / restored Feather nodes when they fell off tip, then decommissioned some RPC nodes from the pool without dropping the public RPC
Public /health no longer hangs if one sub-service is dead
Full mobile UI shipped

What shipped

Mobile: full mobile layout shipped.

Some RPC nodes in the pool lagged in late July. We drained the lagging nodes from the community pool, rerouted traffic, then brought them back in when they caught up.

Cosmos gRPC is forwarded on rpc.gonka.gg:9090 (TLS at the edge, our Feather nodes behind it). Weight / drain when a Feather is in recovery after an upgrade.

Live gateway cut over to a new box. Moved to better servers. ClickHouse + tx scanner follow the archive host. One Feather decommissioned. Pool still serves as it was, just faster and more stable. Health probes are time-bounded so a dead scanner cannot freeze /health or the homepage "Powered by" strip.

Why it matters

Wallets and the explorer only look healthy if the public RPC stays on tip. This stretch was: move the box, add gRPC, do not take Keplr down while we do it. So we did.

Links: Gateway · Docs · Health · gRPC: rpc.gonka.gg:9090 · Access (optional)

7. G-Meter

This month is ops hardening + QA + DeepSeek / OpenBroker / AI status.

Live: https://meter.gonka.gg

What shipped

/compare linked in the top nav. Explained aggregate "failed tests %" vs latest-run broker list (history includes older underpayment failures).

Front QA: Limits / Providers / pricing UI; live /v1/models prices, shared price-scale domain, CIS language on compare. Mobile probe-log + providers layout.

DeepSeek-V4-Flash-0731 added to the active probe lineup (catalog / probes / brokers). Trends filter dedupe. Dropped DeepSeek probes on brokers that do not list it (Hyperfusion / Gonkarouter).

OpenBroker (api.openbroker.gonka.gg) added as OpenBroker by Gonka Labs. DeepSeek via proxy.gonka.gg writes cached network + broker AI summaries (refresh after probes). At a glance opens the broker note from a message-icon modal, with plain-language notes like "billing unavailable".

Why it matters

G-Meter is how anyone can see whether a broker is actually serving, at what price, without trusting a landing page. Adding DeepSeek and our own OpenBroker to that lineup, and not dying during a host cutover, is the point of the page.

Links: Dashboard · Compare · Get started

Still working to make Gonka Labs products better while creating new onsed. Hardening infra under real load, polishing what already shipped, building the next products on the list. More updates as they land.

Links

Feedback welcome in the thread, Gonka Labs team.

Report from Gonka GitHub Discussions

Proposal #74: Gonka Labs - Monthly Report No.1

gonkalabs · Updated 8/5/2026

View on GitHub ↗

This report is community-posted and maintained by the Core team on GitHub, not verified by PTD moderators. Independently confirm important claims.

Hey! Gonka Labs here.

On June 12th the community passed Proposal #74. That vote funded the next stretch of work - infra, ops, and shipping products for the ecosystem.

This is the first public progress report we promised in the proposal (monthly updates on GitHub Discussions). Short version of what landed so far:

OpenBroker - new product, live
Pulse - new product, live
gonka.gg - ~240 QA bugs closed + big product updates (devshards / inference, network analytics, mobile)
proxy.gonka.gg - B2B inference proxy scaled hard + public status/analytics + DevShards v3
Infra move - Whole infra migrated (explorer + indexer stack, all dbs, all backend, new frontend and analytics stack) onto the new production setup.

1. OpenBroker

We already announced the launch here: OpenBroker - broker for brokers or Devshards as a service (Jun 23).

Quick recap: OpenBroker is managed DevShards-as-a-service. Register, deposit GNK, get an obk-* key, point any OpenAI client at https://api.openbroker.gonka.gg/v1 (UI at https://openbroker.gonka.gg). No wallet whitelist dance, no running your own escrow rotation. GNK billing with no markup. Public stats at https://openbroker.gonka.gg/stats.

Since launch (numbers as of Jul 18)

33 active brokers - 48 registered orgs

~3.2GNK deposited into broker ledgers
~15M requests logged in usage (~15M lifetime hits through the dual-gateway router)
~11.8B tokens served (8.5 logged since migration) across MiniMax, Kimi, GLM, Qwen (usage window since early July; launch load test already pushed past 1B tokens in ~1 hour)
Peak observed throughput around ~100 req/s (busiest minutes ~5.7k req/min; busiest day ~1.06M requests)
DevShards v3 live on both production gateways - image mainnet-v0.2.13-v3-post1, route prefix /devshard/v3, 50/50 sticky split behind the router. OpenBroker was already on the v1/v2 path at launch; we completed the network-mandated v3 cutover ahead of the v0.2.14 deadline so brokers do not fall off when classic /v1/devshard goes away
Real brokers / relays already pointing parts or all traffic at OpenBroker (public org names): gonkarelay.com, Gonka24.com, Gonka-API.org, gate.joingonka.ai,Vitarum, etc, plus several private / internal accounts.

Ops + product work after launch

Moved production off the early Railway pilot (where we landed after Qupra DC issues in end of June) onto a dedicated server. Same public domains, more headroom under load
Dual-gateway HA with live traffic drain/switch (used for the v3 migrate - one gateway at a time, verify, then the other)
Broker dashboard: per-request usage, participant / redundancy attempt breakdown, cost reconciliation
New broker APIs: GET /v1/usage and GET /v1/usage/{id} so integrators can pull request-level cost/token data without scraping the UI
Hardening under real load (connection pooling to the gateway router, escrow rotation / capacity ops, model-aware capacity after governance model changes)

Feedback thread (git discussions)

Thanks for the sharp measurements. Where we landed so far:

Billing / hung requests - usage logging + GET /v1/usage/{id} so you can see whether a request settled, what tokens/cost we recorded, and the FinalCost snapshot when the gateway returns it
/v1/models /v1/models /v1/models - still on the polish list (model ids are also in docs / 400 text); not blocking chat completions
~10s fixed overhead - largely protocol / escrow path (block times), not OpenBroker markup. Dual gateways + v3 cutover keep us on the current network path; further shaving depends on protocol-side escrow UX
Large-prompt hang band - treated as a gateway/host-path issue; we keep feeding failure modes back to core while brokers route production traffic through us

OpenBroker is also doing what the launch post promised: shake out new DevShard versions at production scale. v3 on our gateways was exactly that - migrate, smoke /devshard/v3, keep MiniMax brokers online through the cutover.

Links:

Product: https://openbroker.gonka.gg
API: https://api.openbroker.gonka.gg/v1
Register: https://openbroker.gonka.gg/register
Stats: https://openbroker.gonka.gg/stats
Launch post: https://github.com/gonka-ai/gonka/discussions/1363

2. Pulse

Pulse is the live media + sentiment dashboard for Gonka (see it as a backbone and a part of proposal item named MTD / Marketing Transparency Dashboard - we shipped essential part that does data gathering first). It pulls public Gonka coverage from X, Instagram, YouTube, and the web into one place, scores tone via Gonka LLMs (proxy.gonka.gg), and shows activity / reach / engagement next to GNK market data. Built so anyone can see what the ecosystem is saying without opening ten tabs.

Live: https://pulse.gonka.gg

What shipped

Full Pulse app on Railway (API + web + Postgres), collectors on a ~2h cadence
Sources: X, Instagram, YouTube, curated web / press
Dashboard: last-24h activity / reach / engagement charts, general sentiment gauge + 14-day tone chart, GNK market (Uniswap v3 + HEX OTC), market Fear & Greed, top creators, content feed, EN + RU ui translations
Sentiment analysis runs on LLMs running in Gonka (same network we all are building for) - not a closed SaaS LLM
Embedded on the gonka.gg homepage (sentiment, Fear & Greed, latest news, link out to Pulse)
Public read API for stats / market / feed / news / creators so gonka.gg and others can reuse the data. Some dashboards are already ingesting it alongside basic network info.

Numbers (as of Jul 18)

~2,600 posts indexed
~11.9M total reach and ~149K engagement across indexed content
Sentiment: 100% of indexed items analyzed (latest daily reading in the mid-50s / Neutral)
Last 24h snapshot (volatile): ~43 new posts, ~784K reach gained, ~4.5K engagement gained

Why this matters

Proposal #74 called out demand activation and making Gonka easier to follow for non-technical users. Pulse is a part of that "transparency layer" - a public pulse of coverage, tone, and creators, wired into gonka.gg and backed by Networks collective inference stack.

3. gonka.gg - bugs + product work

Explorer side of the proposal was "gonka.gg V2.0" - faster engine, better UI, mobile, accurate GPU/inference metrics. We are not done with that part of roadmap yet, but a lot of the painful day-to-day stuff is already fixed and several new surfaces shipped.

QA bug smash (~240)

We ran a dedicated desktop + mobile QA pass with an external unbiased audit, and closed on the order of ~240 tracker bugs (desktop tracker version into the 170s, mobile into the 60s - still an ongoing flow since we ship new features that are getting QA tested as well).

Rough shape of what that covered:

Mobile layouts and basic UX reorganizations
charts, maps, tooltips escaping the viewport on touch
table, card alignments, repositionings, restructuring
status badges, filters, i18n gaps
reward / weight tooltips and incomplete-epoch edge cases
block / tx / wallet polish
Updated indexers (28k less load on the network)
Devshards V3 flow alignment for inference data ingestion
A lot of backend work to make page load time go down

Shipped across a long streak of QA PRs and multiple levels of sanity checks. Not glamorous, but this is the difference between "works for some people" and "works for 70% of 10k+ people a day on phones".

Product / data updates on gonka.gg (since the vote)

Inference after DevShards (v0.2.12+)

Built and run a dedicated devshard-poller that ingests off-chain session diffs into ClickHouse (the chain no longer emits per-inference txs the old way) and is properly aligned with V3 flow, and is optimized to not make heavy load for the network (overall produced load caused by data ingestion went down X27k).
/network/inference (Inference tab) is driven from that pipeline - 24h stats, gateway traffic, timelines.
Added DevShards v3 support in the poller (v2 and v3 coexist on hosts during migration; we probe /devshard/v3 and union live shard inventories so traffic does not silently drop when gateways switch).

Network analytics

Model Weight Share on /network - weight distribution across AI models per epoch, with an all-epochs view (Redis-cached so it stays fast).
GPU / reward-per-weight chart polish for in-progress epochs.
Live Devshard flow tightened to recent epochs so the UI stays honest under load.

Token holders

Supply fetch fallbacks when LCD nodes flake.
Prune stale holder rows so rankings do not keep wallets that dropped under the qualification threshold.

Infra behind the explorer (also see section 4)

Proper Frontend serving with lowest ttfb possible, API with horizontal auto-scaling, archive ClickHouse + tx-scanner + devshard-poller on dedicated server.
Continuous deploys for explorer fixes without taking the archive down.

Explorer: https://gonka.gg
Inference: https://gonka.gg/network/inference
Network: https://gonka.gg/network

Why it matters

Faster deploys, less risk of a bugfix taking down indexing, room to run the heavier pre-compute work from the V2 explorer roadmap.

4. proxy.gonka.gg - B2B inference + public status

proxy.gonka.gg is our OpenAI-compatible inference proxy on top of Gonka DevShards - API keys, balances, usage, docs, chat playground. Public API at https://api.proxy.gonka.gg/v1. Pulse (and other Gonka Labs apps) call the same stack for LLM work, so when we harden the proxy we harden half the product surface too.

Numbers (as of Jul 18)

Measured window below is since the Jul 4 dedicated-server migrate (usage DB moved with the stack - earlier Railway and even earlier Qupra DC history is not in this series). The product itself was already live before Proposal #74.

~29.0B tokens served (~5.2M requests, ~96.8% OK)
~23.6B tokens in the last 7 days (~3.5M requests)
~7.3B tokens in the last 24 hours (~1.1M requests)
Peak recent days ~4.8-5.0B tokens / day; peak hour observed ~823M tokens in a single hour (~270k requests)
Model mix (token share): MiniMax-M2.7 ~25.6B, Kimi-K2.6 ~3.4B, plus residual Qwen / other
Accounts: +10-30 new accounts daily.

What shipped since the vote

Dedicated production box for the API (api.proxy.gonka.gg) - moved off the flaky VPS / and too unreliable Railway path onto a tuned server with local Postgres, own DevShard gateway, Proxy edge for the API. Frontend stays on other host at proxy.gonka.gg with rewrites into the API.
Own DevShard gateway ops - escrow rotation, capacity-aware concurrency limits, admin tooling. Gateway now on DevShards v3 (mainnet-v0.2.13-v3-post1, route /devshard/v3) ahead of the network v0.2.14 deadline - blue/green cutover with a temp gateway + drain so in-flight MiniMax traffic survived the switch. Currently 32+32 (64 inflight escrows) active v3 escrows (MiniMax + Kimi)
Public status / analytics page at https://proxy.gonka.gg/status - aggregate capacity %, load, in-flight vs cap, active shards, network participant counts, error-rate history + reason breakdown. Deliberately no escrow ids / host IPs (safe to share with customers)
Usage honesty - requested-vs-served model on usage logs when fallback fires, so dashboards show what the client asked for and what actually ran
Product surface kept current for B2B: models page, docs, partner paths, top-up flows, password reset, etc. (dashboard UX work continued alongside the infra move)

Why it matters

Proposal #74 called out proxy.gonka.gg as the B2B / analytics lane. This stretch was less "new landing page" and more "make the pipe carry real load": multi-billion token days, a public status surface customers can trust, and staying on the current DevShard protocol so keys do not die when classic /v1/devshard goes away. Same gateway lessons feed OpenBroker - one network path. More "b2b" work ahead.

Links:

We are still heads-down on the rest of the Proposal #74 checklist - hardening infra under real load, polishing what already shipped, and building the next products on the list. More updates as they land. Stay tuned.

Links

Gonka Labs: https://gonkalabs.com
Explorer: https://gonka.gg
OpenBroker: https://openbroker.gonka.gg
Pulse: https://pulse.gonka.gg
Proxy: https://proxy.gonka.gg
RPC: https://rpc.gonka.gg
GitHub: https://github.com/gonkalabs
OpenBroker launch: https://github.com/gonka-ai/gonka/discussions/1363

Feedback welcome in the thread - what should we prioritize next month from the proposal list?