DeepSeek V4 Pro
Strongest option for algorithmic work, refactors across many files, and problems that need a deliberate reasoning pass.
DeepSeekReasoningCodingavailable
Model ID
Pass this as the model field on every request.
model
deepseek-v4-pro- Provider
- DeepSeek
- Context window
- 256K tokens
- Max output
- 64K tokens
- Access
- Model add-on
- Best for
- Advanced reasoning and software engineering
Endpoints
Both request shapes are accepted for this model.
- /v1/chat/completionsOpenAI chat completions formatavailable
- /v1/messagesAnthropic message formatavailable
Base URL
https://api.prbycode.com/v1At a glance
The numbers most teams compare before switching a route.
Compare with
Other models in the catalog, side by side.
| Model | Context | Input / 1M | Output / 1M | |
|---|---|---|---|---|
| Claude Sonnet 5 | 1M | $3.00 | $15.00 | Compare |
| GLM 5.3 Flash | 256K | $0.25 | $1.10 | Compare |
| GLM 5.2 | 256K | $0.60 | $2.20 | Compare |
| DeepSeek V4.1 Flash | 256K | $0.20 | $0.80 | Compare |
- Claude Sonnet 51M context · $3.00 in / $15.00 out
- GLM 5.3 Flash256K context · $0.25 in / $1.10 out
- GLM 5.2256K context · $0.60 in / $2.20 out
- DeepSeek V4.1 Flash256K context · $0.20 in / $0.80 out