Compare models

Up to four models side by side.

Compare models
Mistral Large 3 Mistral AI Legacy Tiny Aya Global Cohere
Context window 256k, Best in row 8k
Output speed 79.2 No data
Quality 26.7, Best in row 20.1
Reasoning 36.1, Best in row 17.9
Coding 24.7 No data
Vision ✓ −
Tool calling ✓ −
Structured output ✓ −
Self-hostable ✓ ✓
Licence Open weights Open weights
API model id mistral-large-2512 tiny-aya-global

At a glance

Popular comparisons

Head-to-head pages for models people weigh against each other.

Language models 77
Image generation 18
Video generation 11
Text to speech 9
Speech to text 25
Voice and realtime 17
Embeddings 3