Knowledge Base Sections ▾
Navigation
▸ Start here By rolesCategories
- OpenCode: free models, providers and API key
- Cursor + Gonka AI - cheap LLM for coding
- Claude Code + Gonka AI - LLM for the terminal
- DeepSeek Harness: Your Own Provider via JoinGonka Gateway
- Codex CLI: your own key instead of a subscription
- Pi + JoinGonka Gateway — terminal agent on cheap inference
- omp (Oh My Pi) + JoinGonka Gateway: an agent with model roles
- Zoo Code + JoinGonka Gateway: Migrating from Roo Code to Gonka models
- Kilo Code + Gonka AI — AI-Agent in VS Code
- ZCode: GLM Coding Plan tiers, prices and limits
- Hermes Agent + DeepSeek on the Gonka network — an autonomous agent for pennies
- OpenClaw + Gonka AI - affordable AI agents
- Continue.dev + Gonka AI - AI for VS Code/JetBrains
- Cline + Gonka AI - AI agent in VS Code
- Aider + Gonka AI - pair programming with AI
- LangChain + Gonka AI - AI applications for pennies
- n8n + Gonka AI - automation with cheap AI
- Open WebUI + Gonka AI - your own ChatGPT
- LibreChat + Gonka AI — open-source ChatGPT
- Roo Code + Gonka AI — Autonomous AI Agent in VS Code
- LlamaIndex + Gonka AI — RAG applications for pennies
- PydanticAI + Gonka — typed AI agents for pennies
- Vercel AI SDK + Gonka AI — AI applications in TypeScript for pennies
- TanStack AI + Gonka — AI applications in TypeScript for pennies
- API quick start — curl, Python, TypeScript
- JoinGonka Gateway — a full overview
- Management Keys — SaaS on Gonka
- Cheapest AI API: Provider Comparison 2026
- How to buy AI tokens and an API key: 3 methods in 2026
- Free AI API: tokens on sign-up and for referrals
- Cursor Pro request limit reached — breakdown and cheaper alternative
- Cheaper Claude Code: GLM and DeepSeek instead of Claude
- Claude Code + DeepSeek V4 Flash: How to Connect It and What a Task Costs
- Cline is burning money — why the agent spends so much
- OpenClaw is expensive — why the agent burns through tokens and how to save
- OpenRouter alternative: the same models, many times cheaper
- Best AI model for coding in 2026: comparison and prices
- Cheap alternative to GitHub Copilot without limits
- A cheap Windsurf alternative without credits or limits
- The cheapest API for AI agents in 2026
- JetBrains IDE + JoinGonka Gateway — your own endpoint instead of credits
- GitHub Copilot BYOK — own models instead of quotas
- Zed + JoinGonka Gateway — cheap inference in your editor
- MiniMax Code: MiniMax agent with your own key via Gonka
- Warp + JoinGonka Gateway — terminal agent on your own endpoint
- Trae + JoinGonka Gateway — Gonka network models in AI-IDE
- Cherry Studio + JoinGonka Gateway — desktop AI client
- OpenHands + JoinGonka Gateway: agent on your own endpoint
- Qwen Code after the closure of qwen-oauth: working via JoinGonka Gateway
- Goose + JoinGonka Gateway: your own provider and key in the keyring
- Crush + JoinGonka Gateway: Charm agent on Gonka network models
- Kimi Code CLI: Moonshot AI agent on your key via Gonka
- Factory Droid + JoinGonka Gateway: BYOK on Gonka network models
- MiMo Code + JoinGonka Gateway: Xiaomi agent on Gonka network models
Free Start
Free AI API: tokens on sign-up and for referrals
JoinGonka Gateway gives you a free API key to the open models GLM-5.3-Flash, DeepSeek-V4-Flash and MiniMax-M2.7. No bank card needed: after you confirm your email, 3M tokens land in your balance. The key works in any app built for the OpenAI or Anthropic format — in Cursor, Claude Code, n8n or your own bot.
When your starter tokens run out, you can keep going for pennies: $1 buys about 160 million input tokens. You can also earn tokens for free through the partner program: share your link — and a share of the requests from everyone who signs up through it lands in your balance.
What you get for free
- 3M tokens credited to your balance. They arrive after you confirm your email, or immediately if you sign in with Google, GitHub, or Discord. The bonus is one-time, not monthly.
- API key without a card. All you need to sign up is an email. You'll only need a card or crypto if you decide to top up your balance.
- Models and limits as if you paid. Tokens are spent on GLM-5.3-Flash (used by default), DeepSeek-V4-Flash, and MiniMax-M2.7 with the same limit as a paid balance — 120 requests per minute per key.
How far the starter tokens will get you:
| Task | Token usage |
|---|---|
| A question in a chatbot | 1–3K per response |
| Parsing an email or request in n8n automation | 1–5K |
| Summarizing an article or translating a page | 5–15K |
| A coding agent task (Claude Code, Cline) in a small project | from hundreds of thousands |
That's hundreds of chatbot questions or several agent tasks — enough to connect a key and test the models on your own tasks.
How to get a free API key
- Sign up. Open the JoinGonka Gateway registration page, create an account with your email, and confirm it via the link in the email — the tokens will land in your balance. If you sign in with Google, GitHub, or Discord, no email confirmation is needed.
- Key. In the "API keys" section, click "Create key". The key starts with
jg-and is shown in full only once — save it right away. - Connection. In your program, specify the base URL
https://gate.joingonka.ai/v1and your key. Anthropic-format programs need an address without/v1:https://gate.joingonka.ai.
Test the key with your first request — a 200 response means everything is set up:
curl https://gate.joingonka.ai/v1/chat/completions \
-H "Authorization: Bearer jg-your-key" \
-H "Content-Type: application/json" \
-d '{"model": "zai-org/GLM-5.3-Flash", "messages": [{"role": "user", "content": "Say OK"}], "max_tokens": 1024}'For Claude Code, Codex CLI, and OpenCode there's an installer: npx -y @joingonka/setup --tool claude-code will write the settings itself and check the connection. Python and TypeScript examples are in the API quickstart.
When the tokens run out: pricing
After that you only pay for the tokens you use — no subscription and no monthly minimum. One price for all models: $0.0062 per 1M input tokens and $0.019 per 1M output tokens.
The same models cost many times more from other providers. The lowest price on OpenRouter, $ per 1M tokens, input / output, as of November 2, 2026:
| Model | OpenRouter, lowest price | Cheaper with us, input / output |
|---|---|---|
| GLM-5.3-Flash | $0.0285 / $0.0923 | ~5x / ~5x |
| DeepSeek-V4-Flash-0731 | from $0.059 / from $0.14 | ~10x / ~8x |
| MiniMax-M2.7 | $0.21 / $0.84 | ~34x / ~45x |
You can top up your balance with USDT from $1, with the GNK token, or with a bank card — see the article how to buy AI tokens and an API key for details. Payment is upfront, so there won't be a surprise bill at the end of the month.
Tokens for invites: the referral program
Every account has its own referral link: sign up and it will appear in the "Referrals" section of your dashboard. There you can also claim a short address like gate.joingonka.ai/your-name: it's easier to give out in videos and captions.
How it works:
- Someone follows your link and signs up — either right away or within 30 days in the same browser. They get the usual starter tokens and show up in your dashboard in the list of invitees.
- When they use the API, you receive 10% of our commission from each of their requests — about 1% of what they spend.
- If your invitee also invites someone, you receive another 3% commission from those people's requests — about 0.3% of their spend.
- The reward is credited to your balance right away and is spent on the same models.
How much that is in tokens — for first-level invitees:
| Invitees spend per month | You receive per month |
|---|---|
| 100M tokens | about 0.9M tokens |
| 1B tokens | about 9M tokens |
| 10B tokens | about 90M tokens |
For reference: one in ten active Claude Code users through our gateway spends from 20M tokens a day. One such invitee brings you from 4M tokens a month — that's thousands of chatbot questions. So if you use AI lightly, one or two active invitees may be enough to cover your spend entirely.
Where to share the link:
- in an article, post, or video about how you connected Cursor, Claude Code, or n8n to cheap models;
- in your project's README on GitHub or in an automation template that runs through this API;
- in answers in chats and on forums where people ask where to get a free or cheap AI API;
- to colleagues and friends who need a key for their own tasks.
Share where the link fits, without spam. You can't invite yourself: per the service rules, a person can have only one account.
If you need another free option
Other kinds of free AI APIs, terms as of November 2026:
- Permanent free quota — available with the Gemini API in Google AI Studio and with Groq. Requests per minute and per day limits apply, and requests on Google's free tier may be used to improve its products.
- OpenRouter's free models marked
:free— up to 20 requests per minute and 50 per day, and the set of models changes. - A model on your own computer via Ollama or LM Studio — you don't pay for tokens, but models on the level of GLM-5.3-Flash are built for server-grade GPUs.
Frequently asked questions
Do I need a bank card? No. An email is enough to sign up and get your starter tokens.
Is the bonus credited every month? No, it's a one-time bonus: a person can only have one account, so it isn't granted again. You can keep getting tokens through the referral program.
Will the key work for n8n, Cursor or Claude Code? Yes: it works in any program where you can specify your own base URL in OpenAI or Anthropic format. For n8n there's a step-by-step guide.
Where can I see referral rewards? In the "Referrals" section: there you'll find the list of invited users and the accrual history, and the rewards themselves are added to your balance right away.
Does the gateway store my requests? No: the gateway doesn't save request or response texts. The requests themselves are executed by GPU hosts of the decentralized Gonka network.
Your key and starter tokens — free
Sign up with email or via Google, GitHub, Discord — no card needed.
Get free tokens →