PyyanThe AI Index

Pyyan / Compare / Atria Dawn Preview vs DeepSeek V4.1 Flash

Atria Dawn Preview vs DeepSeek V4.1 Flash

2 of 5

Open-Weight Models · verified 16 Sept 2026

SAtria Dawn PreviewShanghai AI Laboratorycurrent
DeepSeek V4.1 FlashDeepSeekcurrent
3 slots left
SpecificationAtria Dawn PreviewDeepSeek V4.1 Flash
Summary744B agentic weights under the MIT licence, published without a launch.Twice the parameters of the Flash it replaces, and cheaper to run.
Context256K1M
Input ($/Mtok)Not published$0.3
Output ($/Mtok)Not published$1.2
LicenceMITMIT
Max outputNot published256K
Parameters744B total552B total, 8B / 16B active
Released14 Sep 202610 Sep 2026
Model IDinternlm/Atria-Dawn-Previewdeepseek-flash
CategoryOpen-Weight ModelsOpen-Weight Models
OfficialShanghai AI Laboratory ↗DeepSeek ↗

Highlighted rows are where these differ.

Atria Dawn Preview

  • Built on Zhipu's open GLM-5.2 rather than trained from scratch
  • 96.0 on DeepSearchQA and 86.5 on CyberGym, both reported by the lab
  • BF16 and FP8 checkpoints, served through SGLang or vLLM

Best for agentic work on your own cluster.

Full spec sheet →

DeepSeek V4.1 Flash

  • 552B total, 8B active while reading the prompt and 16B while generating, with a 1M token context
  • DeepSeek says it beats V4-Pro; V4-Pro requests route to it from 14 September 2026
  • MIT licence, native image input, and a KV cache about a quarter the size of V4-Flash's

Best for cheap agentic coding with images.

Full spec sheet →