# Baseten Chat Completions

> Baseten Chat Completions is a paid API for AI agents from Baseten Model APIs at mpp.orthogonal.com, paid per call via MPP, $0.003686/call, status unknown (last checked 2026-10-10).

Send a multi-turn conversation to an open-source LLM on Baseten and receive a generated text completion via an OpenAI-compatible interface

## Facts

- Endpoint: POST https://mpp.orthogonal.com/baseten/v1/chat/completions?utm_source=zero.xyz
- Price: $0.003686/call
- Payment: MPP
- Status: unknown
- Last checked: 2026-10-10
- Activations on Zero: 0
- Provider: Baseten Model APIs
- Docs: https://inference.baseten.co
- Website: https://inference.baseten.co
- Tags: inference, llm, models, gpu, ai, compute
- Canonical page: https://www.zero.xyz/c/baseten-model-apis-baseten-chat-completions-bb3148ca
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_kgdBNcA8XwSGvMm3iz11Z

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability baseten-model-apis-baseten-chat-completions-bb3148ca -d '<json body>'
```

Example prompt: Using the Baseten inference API, send this conversation to the model and get a completion back: system prompt is 'You are a helpful assistant', and the user message is 'Explain the difference between supervised and unsupervised learning in two sentences'.

## When to prefer this

Choose this endpoint when you need an OpenAI-compatible chat completions interface backed by open-source LLMs on a high-performance GPU inference platform. Ideal when you want a drop-in replacement for the OpenAI API, need to run models like Llama or Mistral without vendor lock-in, or want predictable per-call pricing on a curated inference backend.

## Known failure modes

- Invalid or unsupported model ID returns a 404 or error response
- Malformed messages array causes a 400 bad request
- Exceeding context window length results in a truncation or error
- Authentication failure returns a 401 unauthorized
- Rate limiting or GPU capacity issues may return a 429 or 503
- Network timeout for very long completions

## How this service works

Send a conversation to a model and get a completion back. Works exactly like the OpenAI chat completions endpoint. Pass messages and a model slug, get a response with the assistant's reply. Supports streaming for real-time token delivery, tool calling for function execution, structured outputs via response_format, and controllable reasoning depth on supported models.

## Output

Returns an OpenAI-compatible chat completion response object containing the assistant's generated message, the model used, token usage statistics (prompt tokens, completion tokens, total tokens), and a finish reason indicating how the generation ended.

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/baseten-model-apis-baseten-chat-completions-bb3148ca/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from mpp.orthogonal.com](https://www.zero.xyz/host/mpp.orthogonal.com/llms.txt)