model

Mixtral 8x22B

modelactiveprovisionalmixtral-8x22b-0b1064ea·1 events·first seen 15d ago

Aliases: Mixtral 8x22B

Co-occurring entities

Mistral AI Llama 2 70B Apache 2.0 Sparse Mixture-of-Experts La Plateforme Mixtral 8x7B Mistral 7B HumanEval MMLU GSM8K

More like this (12)

Mixtral 8x7B Instruct Mixtral 8x7B Mixtral-8x7B-v0.1 Mixtral LLaDA-8B Pixtral 12B EvoCUA-8B LLaDA-8B-Base Mistral Compute LLaDA-1.5-8B ms-marco-MiniLM-L-6-v2 DeepSeek-R1-Distill-Llama-8B

Recent events (1)

8Mistral Ai News·15d ago·source ↗

Mistral AI Releases Mixtral 8x22B Under Apache 2.0

Mistral AI has released Mixtral 8x22B, a sparse Mixture-of-Experts model with 141B total parameters but only 39B active parameters, under the permissive Apache 2.0 license. The model features a 64K token context window, native function calling, multilingual support across five European languages, and strong math and coding performance. Mistral claims it outperforms all other open-weight models on standard benchmarks while being faster than dense 70B models due to sparse activation. An instructed version achieves 90.8% on GSM8K maj@8.

Frontier Model Releases Open Weights Progress Mistral AI Llama 2 70B Apache 2.0 +10 more