All times Eastern (ET)Provider pages checked daily at
ProSign inRSS

Mixtral 8x22B

API ID open-mixtral-8x22b

Mixtral 8x22B was a Mistral model. It had a 64,000-token context window.

A large open MoE model competing with much larger dense models.

Mistral's description
Status
Retired
Released
—
Shutdown
Mar 30, 2025

open-mixtral-8x22b stopped serving on March 30, 2025. Recommended replacement: mistral-small-2603.

How to migrate to Mistral Small 4

Benchmarks

Scores as each source published them
or

Comparing with Mixtral 8x7B, the previous model in this line. A score is compared only with the same benchmark from the same source, at the same setting where both have it. Full comparison: prices, limits, features

Arena

ArenaMixtral 8x22BMixtral 8x7BChangeRankDateSource
Text
By category
  • Math 1228 rank 307 vs 1191 (+37)
  • Creative writing 1190 rank 334 vs 1159 (+31)
  • Instruction following 1215 rank 327 vs 1181 (+34)
  • Hard prompts 1244 rank 327 vs 1212 (+32)
  • Long prompts 1220 rank 337 vs 1184 (+36)
  • Multi-turn 1189 rank 339 vs 1168 (+21)
  • Expert 1219 rank 322 vs 1197 (+22, within range)
1230 ±51197 ±4+33331 of 415
51,416 votes		Oct 13, 2026source
Text: Coding1278 ±91240 ±8+38322 of 410
8,780 votes		Oct 13, 2026source

Rated as mixtral-8x22b-instruct-v0.1; Mixtral 8x7B as mixtral-8x7b-instruct-v0.1. Ratings come from people voting between two anonymous answers; Text is style-controlled, as on arena.ai. Data: Arena (arena.ai), CC BY 4.0.

Epoch AI

BenchmarkMixtral 8x22BMixtral 8x7BChangeSettingRunSource
GPQA Diamond34.1%30.6%+3.5 ptsJan 27, 2025source
MATH Level 524.2%9.3%+14.9 ptsJan 27, 2025source

Run independently by Epoch AI. Data: Epoch AI, Capabilities & benchmarking (epoch.ai/benchmarks), CC BY.

Features

Compare models
FeatureValueSinceSource
Context windowThe most the model can work with in one request, in tokens: your prompt, conversation history and files. A token is about three-quarters of an English word.64,000 tokensin place when tracking began (October 6, 2026)source

Not published by Mistral: Structured outputs, Function calling, Prompt caching, Batch API, Image input, PDF input, Reasoning / thinking, Fine-tuning, Built-in web search, temperature parameter, top_p parameter, logprobs, Max output tokens.

Price history

USD per 1M tokens, standard tier

No price history yet.

Prices

USD per 1M tokens, every type and band, by tier

No published prices.

Changelog

Changes we detected on Mistral's pages

No changes recorded yet.

Last checked · Sources: Mistral models