# mistralai/mistral-inference

Official inference library for Mistral models

Repository: https://github.com/mistralai/mistral-inference
Canonical: https://ross.abutalabs.com/products/mistral-inference
Homepage: https://mistral.ai/
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: llm, llm-inference, mistralai
Archived: true
Last push: 2026-06-16T15:40:44+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 87, release rhythm 8, longevity 76
- inputs: {"age_days": 1071, "days_push": 78, "days_rel": 531, "gap_med": 188, "n_releases_24m": 2}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 10832, forks 1056 (observed 2026-08-28T04:10:44.446933+00:00)

## What it is
The official minimal Python library from Mistral AI for running inference with Mistral open-weight models such as Mistral 7B, Mixtral, Codestral, and Pixtral locally on GPUs. The repository is now archived and no longer actively maintained.

## Use cases
- run mistral 7b locally on my gpu
- generate text with mixtral 8x7b
- run codestral for code completion locally
- chat with a mistral instruct model from python
- run pixtral vision model locally
- self-host mistral open weights

## When to choose
- you need the official minimal reference implementation for running Mistral open-weight models
- you want a lightweight local inference setup without a heavy serving framework
- you are reproducing tutorials or research based on Mistral model weights

## When to avoid
- you need actively maintained software with bug fixes and new model support
- you want production serving with batching, quantization, or an OpenAI-compatible API (use vLLM or similar)
- you have no GPU, since installation requires xformers and a GPU
- you want to use the Mistral cloud API instead of local weights

## Facets
- artifact type: library
- maturity: abandoned
- function: llm-inference, machine-learning
- domain: large-language-models, machine-learning, artificial-intelligence
- platform: python
- tags: mistral, open-weights, transformers, archived, chat-models, gpu, linux

## Member repositories
- mistralai/mistral-inference (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:44.446933+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:17:38.149666+00:00, confidence not recorded.
  - readme: https://github.com/mistralai/mistral-inference (fetched 2026-08-28T04:10:44.446933+00:00, sha d3102c41556e)
  - homepage: https://mistral.ai/ (fetched 2026-08-29T08:16:37.798582+00:00, sha f0b4c411ac9d)
  - site_page: https://docs.mistral.ai (fetched 2026-08-29T08:16:37.809276+00:00, sha e616637c1e71)
  - site_page: https://mistral.ai/about (fetched 2026-08-29T08:16:37.811254+00:00, sha de97b67551c6)
  - site_page: https://mistral.ai/pricing (fetched 2026-08-29T08:16:37.801304+00:00, sha eff2324488f8)
  - site_page: https://mistral.ai/pricing/api (fetched 2026-08-29T08:16:37.803384+00:00, sha 246e18e81945)
  - site_page: https://mistral.ai/pricing/enterprise-deployments (fetched 2026-08-29T08:16:37.806870+00:00, sha a067af91b924)
- Data as of 2026-08-30T08:39:29.467469+00:00.