LLM API — OpenAI-compatible
Claude, GPT, Gemini, Grok, DeepSeek, Kimi and more behind one OpenAI-compatible /v1/chat/completions endpoint. Keep your SDK, change the base URL, pay per call.
35 models from 11 providers: Anthropic, OpenAI, Google, xAI, MiniMax, DeepSeek, Thinking Machines, Moonshot AI, Alibaba, TypeSafe, Unbiased.
Models
| Model | Provider | Price |
|---|---|---|
| claude-fable-5 | Anthropic | $0.0538 Per call (1k-token answer) |
| claude-opus-4.8 | Anthropic | $0.0269 Per call (1k-token answer) |
| gpt-5.5-pro | OpenAI | $0.1936 Per call (1k-token answer) |
| gpt-5.5 | OpenAI | $0.0323 Per call (1k-token answer) |
| gemini-3.1-pro | $0.0129 Per call (1k-token answer) | |
| grok-4.3 | xAI | $0.002691 Per call (1k-token answer) |
| minimax-m3 | MiniMax | $0.001292 Per call (1k-token answer) |
| deepseek-v4-pro | DeepSeek | $0.003745 Per call (1k-token answer) |
| inkling | Thinking Machines | $0.005160 Per call (1k-token answer) |
| inkling-256k | Thinking Machines | $0.0103 Per call (1k-token answer) |
| gemini-3.6-flash | $0.008073 Per call (1k-token answer) | |
| gemini-3.5-flash-lite | $0.002665 Per call (1k-token answer) | |
| grok-4.5 | xAI | $0.006563 Per call (1k-token answer) |
| gpt-5.6-luna | OpenAI | $0.006432 Per call (1k-token answer) |
| gpt-5.6-terra | OpenAI | $0.0161 Per call (1k-token answer) |
| kimi-k3 | Moonshot AI | $0.0161 Per call (1k-token answer) |
| claude-sonnet-5 | Anthropic | $0.0161 Per call (1k-token answer) |
| claude-opus-5 | Anthropic | $0.0269 Per call (1k-token answer) |
| gpt-5.6-sol | OpenAI | $0.0322 Per call (1k-token answer) |
| qwen3.7-plus | Alibaba | $0.001386 Per call (1k-token answer) |
| qwen3.7-max | Alibaba | $0.004102 Per call (1k-token answer) |
| qwen3.8-max | Alibaba | $0.006563 Per call (1k-token answer) |
| grok-4.6 | xAI | $0.006563 Per call (1k-token answer) |
| gemini-3.7-flash | $0.004037 Per call (1k-token answer) | |
| gemini-3.8-flash | $0.004037 Per call (1k-token answer) | |
| claude-fable-5.1 | Anthropic | $0.0538 Per call (1k-token answer) |
| gpt-6-astra | OpenAI | $0.0551 Per call (1k-token answer) |
| grok-4.7 | xAI | $0.006714 Per call (1k-token answer) |
| jev | TypeSafe | $0.001412 Per call (32k-token ceiling) |
| pareto | Unbiased | $0.008393 Per call (1k-token answer) |
| claude-opus-5.5 | Anthropic | $0.0215 Per call (1k-token answer) |
| claude-sonnet-5.5 | Anthropic | $0.0108 Per call (1k-token answer) |
| gpt-6-luna | OpenAI | $0.000539 Per call (1k-token answer) |
| gpt-6-sol | OpenAI | $0.0108 Per call (1k-token answer) |
| gpt-6.1-sol | OpenAI | $0.0108 Per call (1k-token answer) |
Call it
The first response is 402 with the exact price and every way to pay. Pay,
repeat the request with the proof, and you get the result.
curl -X POST https://api.glianalabs.com/v1/chat/completions \
-H 'content-type: application/json' \
-d '{"model":"gpt-5.5","messages":[{"role":"user","content":"Summarise x402 in one sentence."}],"max_tokens":256}'How it works
The request and response follow OpenAI chat completions, including streaming (SSE), tool calling and image inputs on vision models. Pick the model per request. The same endpoint is also reachable over x402 at /x402/chat/completions.
Limits worth knowing
A chat call is paid before the model runs, so it is priced at its ceiling: the estimated input plus max_tokens of output. Set max_tokens to what you need; a large or missing max_tokens means a large quote.
Questions
Can I use the OpenAI SDK?
Yes. Set base_url to https://api.glianalabs.com/v1 and wrap the SDK’s fetch with a payment client such as mppx, or pass a prepaid gk_ key as the API key.
How are tokens billed?
Up front, at the model’s rates, for the estimated input plus max_tokens. That is the most the call can cost, so it is never undercharged mid-stream.
Are tool calls and streaming supported?
Yes, on models that support them, in the OpenAI format.
Is there a free tier?
No. Every call is paid individually, with no plan or minimum. A request that fails validation is refused before it runs, so it is never charged.
Do I need an account or an API key?
No. The first response is HTTP 402 with the exact price; pay it with USDC from your own wallet and repeat the request. To prepay instead, top up a balance in rupiah and send a gk_ key.