DeepSeek V4.1 Flash
ModelOpen sourceActiveby DeepSeek · family “deepseek-v4-flash”
DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on... (description from the OpenRouter listing)
in: textin: imageout: textReasoningTool useStructured output
Epoch Capabilities Index
155.0 #28 of 268
90% CI 148.5 – 157.5
Released
9 Sep 2026
Context window
1.05M tokens
Max output
384K tokens
Input $ / 1M tokens
$0.035
Output $ / 1M tokens
$0.29
Cached input $ / 1M
$0.001
Each benchmark shown separately with its own source
Benchmark results (1)
| Benchmark | Domain | Score | vs best recorded | Setting | Run | Source |
|---|---|---|---|---|---|---|
| Surface Evolver Bench | science | 46.3% | 49% | high | — | External ↗ |
Source: Epoch AI Benchmarking Hub (CC BY 4.0). “External” rows are leaderboard results Epoch collects from third parties. Best reported setting per benchmark is shown.
API list price over time
Price history
$ per 1M tokens
No data recorded for this period
1 observation (OpenRouter listing). History accumulates with every ingest run; a single point means no change has been observed yet.
Events
- Open release20d agoDeepSeek releases DeepSeek V4.1 FlashEpoch AI Benchmarking Hub