AICreatorHub
NewsToolsModelsGuides
AICreatorHub

India's bilingual AI knowledge hub.

ExploreNewsToolsModelsGuides
LegalAboutContactPrivacy PolicyTermsDisclaimer
FollowX / TwitterYouTubeRSS
© 2026 AICreatorHub. All rights reserved.
HomeModelsMistral Large

Mistral Large

Mistral AITEXT Open weight
Provider
Mistral AI
Modality
TEXT
Parameters
123B
Context window
128,000 tokens
Weights
Open
Released
—

Mistral Large is Mistral's flagship model, balancing strong quality with efficiency, with open-weight options that make it self-hostable.

It is popular for multilingual chat, coding and on-prem deployments in Europe and beyond.

Best for

  • Multilingual chat and content
  • Efficient self-hosted deployments
  • Coding and structured output

Strengths & how it compares

  • Strong quality-to-efficiency ratio.
  • European provider with clear licensing.
  • Vs Llama 3.3 70B: comparable general quality; vs Qwen: weaker Hindi but strong European languages.

Benchmarks

Representative public scores (approximate, higher is better) for relative comparison. Check the provider for the latest official results.

MMLU (knowledge)84
GPQA (reasoning)48
HumanEval (coding)78

How to access

Open weights for self-host; managed API per token.

Access Mistral Large
Share: