deepseek-reasoner
deepseek-reasoner now serves DeepSeek-R1; see official R1 release notes.
DeepSeek's change log says: deepseek-reasoner is our new model DeepSeek-R1. You can invoke DeepSeek-R1 by specifying model='deepseek-reasoner'. For details, please refer to: DeepSeek-R1 Release
Who it affects: API users calling model=deepseek-reasoner.
What it means: deepseek-reasoner now serves DeepSeek-R1; see official R1 release notes.
Official sources
Related pages: /guides/model-name-migration · /models/deepseek-r1 · /news
Context for developers
Before this entry, deepseek-reasoner was a distinct product string. Afterward it mapped to DeepSeek-R1 weights documented in the DeepSeek-R1 README (GRPO training, visible chain-of-thought). API users selected model="deepseek-reasoner" in OpenAI-compatible calls.
R1 remains available as open weights and in research write-ups; the commercial API table today emphasizes V4.1 Flash and V4 Pro. Compare behavior on the R1 model page and official change log rather than cached blog posts.
Independent resource, not affiliated with DeepSeek.