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×

o3×

Gemini 3 Pro×Clear all

	Gemini 3.5 Google	Llama 4 Maverick Meta	o3 OpenAI	Gemini 3 Pro Google
Vendor	Google	Meta	OpenAI	Google
Family	Gemini	Llama	o-series	Gemini
Release date	2026-05-20	2025-04-05	2025-04-16	2026-04-22
Context window	2,000,000 tokens	1,000,000 tokens	200,000 tokens	2,000,000 tokens
Parameters	—	400B	—	—
Modality	text, vision, audio, video	text, vision	text, vision	text, vision, audio, video
License	proprietary	Llama 4 Community License	proprietary	proprietary
Source	proprietary	open weights	proprietary	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.	Reasoning-focused model in the o-series.	Google DeepMind's flagship multimodal model. Best-in-class for multimodal understanding and agentic / vibe coding workflows. A "Deep Think" reasoning mode is available to AI Ultra subscribers.
Links	vendor →	weights →		vendor →
Benchmarks
MMLU-Pro	88.9%	80.5%	—	87.1%
GPQA-D	91.2%	—	87.7%	88.6%
HumanEval	—	—	—	96.5%
Aider	—	—	—	84.8%
AIME-25	96.1%	—	88.9%	92.8%
LiveCB	—	—	—	76.1%
MMMU	86.0%	68.5%	—	84.3%