# Mistral Large 3

Long-context assistants and enterprise agents. Mistral's 675B granular MoE (41B active), for long-context assistants and enterprise agents. Served on GraphN managed inference behind an OpenAI-compatible API as `mistral-large-3`.

## Overview

Mistral Large 3 is a general-purpose granular mixture-of-experts model with 675B total and 41B active parameters, post-trained for chat, agentic and instruction-following work. Mistral designs it for production assistants, retrieval-augmented systems and enterprise workflows, with native function calling, strong system-prompt adherence, a 256K context window and support for dozens of languages.

## Strengths

- Native function calling and JSON output
- 256K context window
- Dozens of supported languages

## Use cases

- Retrieval-augmented assistants over long documents
- Enterprise knowledge work and agents
- General coding assistance

## Good to know

The weights include a vision encoder, but image input is not honoured on this alias, so treat it as text-only.

## Specifications

- **Type**: Chat
- **Context**: 256K context
- **Parameters**: 675B total, 41B active (MoE)
- **License**: Apache-2.0
- **Family**: Mistral Large 3
- **Weights**: mistralai/Mistral-Large-3-675B-Instruct-2512
- **Pricing**: Input $0.50/1M · Output $1.50/1M

## API access

### curl

```bash
curl https://model.graphn.ai/v1/chat/completions \
  -H "Authorization: Bearer $GRAPHN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistral-large-3",
    "messages": [{"role": "user", "content": "Hello"}]
  }'
```

### CLI

```bash
graphn model chat mistral-large-3 -m "What is GraphN?"
```

### Python

```python
@function
async def ask(question: str) -> str:
    return await chat.complete(question, model="mistral-large-3")
```

[Start free](https://graphn.ai/signup?next=%2Finference%2Fmistral-large-3)

[Model card](https://huggingface.co/mistralai/Mistral-Large-3-675B-Instruct-2512) · [Managed inference](https://graphn.ai/inference) · [Pricing](https://graphn.ai/pricing) · [API reference](https://graphn.ai/api)