DeepSeek-V4-Flash
DeepSeekLanguage modeling/generationQuestion answeringOpen weightsmit
DeepSeek-V4-Flash is a language modeling/generation model from DeepSeek released in 2026 with 284000000000.0 parameters. It is available through the transformers library and is released under the mit license, and has 1.1M downloads and 5.1K likes.
About DeepSeek-V4-Flash
We present a preview version of DeepSeek-V4 series, including two strong Mixture-ofExperts (MoE) language models — DeepSeek-V4-Pro with 1.6T parameters (49B activated) and
DeepSeek-V4-Flash with 284B parameters (13B activated) — both supporting a con
LLM pricing & performance
Full LLM page →DeepSeek-V4-Flash is available via API — live cost, context, and benchmark data:
Input / 1M
$0.14
Output / 1M
$0.28
Context
1M
Tokens/sec
—
Details
- Provider
- DeepSeek
- Task
- Language modeling/generation,Question answering
- Parameters
- 284000000000.0
- Library
- transformers
- License
- mit
- Released
- 2026-04-24
- Open weights
- Yes
- Downloads
- 1107211
- Likes
- 5101