# Mistral Saba API

Mistral's Mistral Saba, through an OpenAI-compatible API. Providers quote it on an open exchange, and each request goes to the offer with the lowest estimated charge that meets your rules.

- Providers compete for each request. The exchange ranks the offers that meet your rules by estimated charge. It fixes your maximum charge before the first token, and you pay for the tokens that the model uses.
- The OpenAI and Anthropic APIs. Use the OpenAI or Anthropic SDK that you have. Set the base URL and your key, and send mistral-saba as the model.
- Your data rules. Send x-liquid-retention: none and only offers whose provider does not keep your content compete for the request.

## Facts

- Model name: mistral-saba
- Maker: Mistral
- Context window: 32.8K tokens
- Input price: $0.199 per million tokens
- Output price: $0.595 per million tokens
- Providers: 1 independent provider
- Regions: us

## Providers

- [Architect](https://liquidinference.ai/providers/805445a7-0e40-4c1b-9baf-7b3b01e30e42.md): us, $0.199 input and $0.595 output per million tokens, $0.199 cache read per million tokens, 32.8K tokens of context

## Scores

- HLE: 4.3, #219 of 273
- GPQA: 42.4, #226 of 250
- MMLU-Pro: 61.1, #116 of 133

Scores: Artificial Analysis (https://artificialanalysis.ai/)

## Request

```sh
curl https://router.inference.ai.exchange/v1/chat/completions \
  -H "Authorization: Bearer $LIQUID_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistral-saba",
    "messages": [{"role":"user","content":"Say hello."}]
  }'
```

## Questions

### What is Mistral Saba?

Served by intent-router: each job goes to the cheapest healthy upstream for this model.

### How much does Mistral Saba cost?

Offers now ask $0.199 per million input tokens and $0.595 per million output tokens. A job pays for the tokens it uses. The exchange picks the cheapest eligible offer.

### What is the context length of Mistral Saba?

The largest offer takes 32,768 input tokens. A smaller offer can take less. The exchange skips an offer that cannot take your message.

### What can Mistral Saba do?

Its providers list Reasoning effort, JSON object, JSON schema, Frequency penalty, Logit bias, Log probabilities, Min-p sampling, Presence penalty, Repetition penalty, Seed, Stop sequences, Top-k sampling, Top-p sampling, Text, Tools.

### What inputs and outputs does Mistral Saba support?

Text → Text.

### Which providers serve Mistral Saba?

1 independent provider quote it now, with 1 offer. The table above ranks them by the cost of a typical job.

### How do I call Mistral Saba?

Send "mistral-saba" as the model in an OpenAI or Anthropic request. The API section links to setup steps for each client.

### When was Mistral Saba listed?

The exchange first listed it on 10/8/2026. This is not its release date.

## Related pages

- [This page in a browser](https://liquidinference.ai/models/mistral-saba)
- [Mistral: Small 3.2 24B Instruct](https://liquidinference.ai/models/mistral-small-3.2-24b-instruct.md)
- [Devstral: Small 2505](https://liquidinference.ai/models/devstral-small-2505.md)
- [Ministral 3B 2512](https://liquidinference.ai/models/ministral-3b-2512.md)
- [Mistral: Nemo Instruct 2407](https://liquidinference.ai/models/mistral-nemo-instruct-2407.md)
- [Cydonia 24B V2](https://liquidinference.ai/models/cydonia-24b-v2.md)
- [Magidonia 24B V4.3](https://liquidinference.ai/models/magidonia-24b-v4.3.md)
- [Set up a client](https://liquidinference.ai/connect/mistral-saba)
- [All models](https://liquidinference.ai/models)
- [Providers](https://liquidinference.ai/providers.md)
- [Discover the best model for your budget](https://liquidinference.ai/discover)
- [Quickstart](https://liquidinference.ai/docs/quickstart.md)
- [llms.txt: the index of every page, as Markdown](https://liquidinference.ai/llms.txt)
- [llms-full.txt: every page in one file](https://liquidinference.ai/llms-full.txt)