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

Gemini 3.5×

Llama 4 Maverick×

DeepSeek-V4-Pro×

Grok 4×Clear all

	Gemini 3.5 Google	Llama 4 Maverick Meta	DeepSeek-V4-Pro DeepSeek	Grok 4 xAI
Vendor	Google	Meta	DeepSeek	xAI
Family	Gemini	Llama	DeepSeek	Grok
Release date	2026-05-20	2025-04-05	2026-04-22	2025-07-09
Context window	2,000,000 tokens	1,000,000 tokens	1,000,000 tokens	256,000 tokens
Parameters	—	400B	1.6T (49B active)	—
Modality	text, vision, audio, video	text, vision	text	text
License	proprietary	Llama 4 Community License	MIT	proprietary
Source	proprietary	open weights	open weights	proprietary
Description	Google DeepMind's next-gen Gemini — positioned as "frontier intelligence with action". Built for complex agentic workflows. Announced at Google I/O 2026.	Mixture-of-experts open-weight model from Meta.	DeepSeek's flagship open-weight MoE. 1.6T parameters with 49B activated, 1M-token context, and a hybrid attention scheme (CSA + HCA) that delivers long-context inference at ~27% of V3.2's FLOPs.	—
Links	vendor →	weights →	weights →	vendor →
Benchmarks
MMLU-Pro	88.9%	80.5%	84.2%	82.8%
GPQA-D	91.2%	—	82.4%	87.5%
HumanEval	—	—	95.1%	—
Aider	—	—	80.1%	72.5%
AIME-25	96.1%	—	88.6%	86.1%
LiveCB	—	—	72.4%	65.3%
MMMU	86.0%	68.5%	—	—