Compare models

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

Selected

Grok 4×

Llama 4 Scout×

Qwen3.6-27B×

DeepSeek-V4-Pro×Clear all

	Grok 4 xAI	Llama 4 Scout Meta	Qwen3.6-27B Alibaba	DeepSeek-V4-Pro DeepSeek
Vendor	xAI	Meta	Alibaba	DeepSeek
Family	Grok	Llama	Qwen	DeepSeek
Release date	2025-07-09	2025-04-05	2026-04-22	2026-04-22
Context window	256,000 tokens	10,000,000 tokens	262,144 tokens	1,000,000 tokens
Parameters	—	109B	27B (dense)	1.6T (49B active)
Modality	text	text, vision	text, vision, video	text
License	proprietary	Llama 4 Community License	Apache-2.0	MIT
Source	proprietary	open weights	open weights	open weights
Description	—	—	Alibaba's first dense open-weight in the 3.6 family. Strong agentic-coding scores (77.2 SWE-bench Verified, matching Claude 4.5 Opus on Terminal-Bench 2.0). Supports 201 languages and multimodal text/image/video input.	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 →	weights →