gpt.buzz
Sign in

Compare models

Pick up to 4models. Specs render side-by-side. Share the URL — it's stateless.

SelectedMistral logoMistral Large 2×DeepSeek logoDeepSeek-V3.1×Google logoGemini 2.5 Flash×DeepSeek logoDeepSeek-R1×Clear all
 
Mistral logoMistral Large 2

Mistral

DeepSeek logoDeepSeek-V3.1

DeepSeek

Google logoGemini 2.5 Flash

Google

DeepSeek logoDeepSeek-R1

DeepSeek

VendorMistralDeepSeekGoogleDeepSeek
FamilyMistralDeepSeekGeminiDeepSeek
Release date2024-07-242025-08-212025-01-20
Context window128,000 tokens128,000 tokens1,000,000 tokens128,000 tokens
Parameters123B671B671B
Modalitytexttexttext, vision, audiotext
LicenseMistral Research LicenseMITproprietaryMIT
Sourceopen weightsopen weightsproprietaryopen weights
DescriptionLarge MoE open-weight model. Predecessor to DeepSeek-V4.Reasoning-focused open-weight model.
Links

Add a model

Max 4 models. Remove one to add another.