DeepSeek V4 Pro
DeepSeek’s large open MoE, GA release (0813).
DeepSeek V4 Pro 0813 is the GA release of DeepSeek’s large-scale mixture-of-experts model. Performance figures below are for the non-reasoning mode.
- Provider
- DeepSeek
- Type
- Language model
- Released
- Aug 12, 2026
- Context window
- 1M tokens
- Max output
- 393K tokens
- Price
- $0.66 input / $1.98 output per 1M tokens
- Input
- text
- Output
- text
- Open weights
- Yes
- License
- MIT
Best for
- Chat & assistants
- Coding
- High volume / low cost
- Local & on-device
Strengths
- Low price
- Open weights
More from DeepSeek
- DeepSeek V4.1 Flash — DeepSeek. Open-weight, very fast MoE built on DeepSeek’s new encoder-decoder architecture.