Compare models

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

Comparing agents instead? Switch to agent compare →

Selected

Llama 4 Maverick×

DeepSeek-V3.1×

Mistral Large 2×

Grok 4×Clear all

	Llama 4 Maverick Meta	DeepSeek-V3.1 DeepSeek	Mistral Large 2 Mistral	Grok 4 xAI
Vendor	Meta	DeepSeek	Mistral	xAI
Family	Llama	DeepSeek	Mistral	Grok
Release date	2025-04-05	2025-08-21	2024-07-24	2025-07-09
Context window	1,000,000 tokens	128,000 tokens	128,000 tokens	256,000 tokens
Parameters	400B	671B	123B	—
Modality	text, vision	text	text	text
License	Llama 4 Community License	MIT	Mistral Research License	proprietary
Source	open weights	open weights	open weights	proprietary
Description	Mixture-of-experts open-weight model from Meta.	Large MoE open-weight model. Predecessor to DeepSeek-V4.	Large text-only Mistral model with a 128K context window and 123B parameters, tuned for strong instruction following and long-context reasoning. Mistral's flagship open-weights release in the Large 2 line.	—
Links	weights →	weights →	vendor →weights →	vendor →
Benchmarks
MMLU-Pro	80.5%	—	—	82.8%
GPQA-D	—	—	—	87.5%
HumanEval	—	—	89.0%	—
Aider	—	—	—	72.5%
AIME-25	—	—	—	86.1%
LiveCB	—	—	—	65.3%
MMMU	68.5%	—	—	—