DeepSeek
DeepSeek Flash
deepseek/deepseek-flash- Context
- 1M tokens
- Token rate
- 1.0×
- Speed
- Fast
- Hosting
- Direct API
Maximum prompt length
Relative to baseline
Estimated output rate
Optimised routing
Description
DeepSeek-V4.1-Flash — 1M context, thinking (default) + non-thinking, vision, up to 384K output
Modalities
- Text in / out
- Supported
- Image / vision input
- Supported
- Audio in / out
- Not supported
- Video synthesis
- Not supported
Pricing
- Baseline per token
- ~$0.0000033
- Adjusted rate
- 1.0× baseline
- Included in
- All paid plans