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

GPT-5×

Gemini 3.5 Flash×

DeepSeek-V4-Pro×

GPT-5 Pro×Clear all

	GPT-5 OpenAI	Gemini 3.5 Flash Google	DeepSeek-V4-Pro DeepSeek	GPT-5 Pro OpenAI
Vendor	OpenAI	Google	DeepSeek	OpenAI
Family	GPT	Gemini	DeepSeek	GPT
Release date	2025-08-07	2026-05-20	2026-04-22	2025-08-07
Context window	400,000 tokens	—	1,000,000 tokens	400,000 tokens
Parameters	—	—	1.6T (49B active)	—
Modality	text, vision, audio	text, vision, audio	text	text, vision, audio
License	proprietary	proprietary	MIT	proprietary
Source	proprietary	proprietary	open weights	proprietary
Description	OpenAI's flagship general-purpose model. Superseded by GPT-5.5 in April 2026.	Fast, low-cost tier of the Gemini 3.5 family announced at Google I/O 2026. June 2026 update added native computer-use capability, moving Flash from generation to agentic task execution.	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.	Higher-reasoning variant of GPT-5, used for harder scientific and research tasks (e.g., immunology breakthroughs at NIH). Same 400K context as GPT-5; trades latency for depth.
Links	vendor →	vendor →	weights →	vendor →
Benchmarks
MMLU-Pro	85.1%	—	84.2%	—
GPQA-D	87.3%	—	82.4%	—
HumanEval	95.2%	—	95.1%	—
Aider	85.4%	—	80.1%	—
AIME-25	91.0%	—	88.6%	—
LiveCB	74.5%	—	72.4%	—
MMMU	74.1%	—	—	—