available
Weights
Official weights are downloadable.
Inspect this claim’s source ↗DeepSeek / FAMILY RECORD
DeepSeek V4.1 Flash
A downloadable DeepSeek multimodal Flash model, with sparse activation that differs between prompt processing and generation.
Checked release: DeepSeek-V4.1-Flash
Sources checked 2026-09-19
- DeepSeek API availability and new pricing
- · Primary source
The cited event establishes API availability and pricing on this date; it does not independently timestamp the weight upload.
The profile describes this checkpoint, not every release carrying the family name. Developer documentation is not independent reproduction.
THE MODEL / THE TERMS
Know what you are picking up.
- Parameters
- 552B backbone; 8B active in prefill / 16B in decode. Hugging Face displays 763B in model files; no single total-storage estimate is asserted.
- Architecture
- Sparse mixture-of-experts multimodal transformer; 8B active during prefill and 16B during decoding.
- Modalities
- text · image
- Context
- 1M tokens
- Weights license
- MIT
- Commercial use
- Permitted by the MIT license.
- Memory
- No reliable universal minimum is stated; the very large checkpoint requires distributed or aggressively quantized serving.
- Known quantizations
- Official and community formats should be verified against the model card before use
Specifications are developer-reported. Primary release documentation ↗ · All sources and terms
NINE DIMENSIONS / NO OPENNESS SCORE
What is actually available?
Each component has its own evidence. “Not established” means this source review did not verify availability; it is not a claim that nothing exists.
available
Architecture
The model card documents the sparse architecture and activation behavior.
Inspect this claim’s source ↗available
Inference code
Official prompt encoding and inference examples are provided.
Inspect this claim’s source ↗not established
Training code
Complete pre-training and post-training code was not confirmed.
Inspect this claim’s source ↗partial
Training recipe
The release explains major techniques but is not an end-to-end reproducible recipe.
Inspect this claim’s source ↗partial
Data information
Some training approach information is disclosed without a complete corpus inventory.
Inspect this claim’s source ↗not established
Training data
No sufficiently specific public artifact was confirmed in this review.
Inspect this claim’s source ↗available
Evaluation
Evaluation reproduction instructions are included with the release.
Inspect this claim’s source ↗available
Commercial use
The official checkpoint is MIT-licensed.
Inspect this claim’s source ↗WHY THIS RELEASE MATTERS
What to explore.
- Long context
- Multimodal input
- Sparse inference
WHAT THE RECORD DOES NOT ESTABLISH
Keep the limits.
- Displayed file size and stated backbone parameter count differ
- Training corpus is not released
No independent evaluation is attached to this record.
FOLLOW THE RECORD
Sources, not superlatives.
- DeepSeek V4.1 Flash announcement ↗ (institution)
- DeepSeek-V4.1-Flash model card ↗ (registry)
Source check: 2026-09-19. AI-assisted editorial review, with no named external expert or independent hardware validation claimed. Licenses and model cards can change. Review the linked terms for the specific artifact and intended use.
Understand open source vs open weights ↗