DeepSeek · open weights
DeepSeek V4.1-Flash
The native text/image-to-text checkpoint is published under MIT, with a 552B backbone, 196B Engram conditional memory and a native one-million-token context label. September 10 is the hosted launch date; the exact native checkpoint publication day is unverified.
- Organisation
- DeepSeek
- Weights published
- Not verified
- Licence
- MIT
- Size
- 552B MoE backbone + 196B Engram conditional memory; 8B active prefill / 16B decode
- Modalities
- textimage
- Comparison catalog
- DeepSeek V4.1 Flash profile
Practical fit
Check actual model files, mixed precision, conditional-memory placement, cache and the chosen backend. The backbone count alone omits Engram storage; a generic uniform-Q4 calculation does not establish native pack size or desktop fit. The 8B/16B active counts are not total weight-memory requirements.
Deployment note
The repo now publishes a minimal inference reference with vision, Engram, MoE and DSpark paths; generation is plain autoregressive sampling. Its self-test uses uninitialized weights and checks shapes/kernel plumbing, not real model behaviour. Native effort is 1–100 with low=50/high=75/max=100 and default high=75; this does not establish the hosted API default or output cap. Production serving and an actual local install were not tested.