3 ms·
> max_tokens：The maximum length of the final response after the CoT output is completed, defaulting to 4K, with a maximum of 8K. Note that the CoT output can re
by throwup238 2y ago
> max_tokens：The maximum length of the final response after the CoT output is completed, defaulting to 4K, with a maximum of 8K. Note that the CoT output can reach up to 32K tokens, and the parameter to control the CoT length (reasoning_effort) will be available soon. [1]
[1] https://api-docs.deepseek.com/guides/reasoning_model https://api-docs.deepseek.com/guides/reasoning_model
- gliptic 2y agoSo yes, it's a limitation of their own API at the moment, not a model limitation.
- throwup238 2y agoI’m using it through Kagi which doesn’t use Deepseek’s official API [1]. That limitation from the docs seems to be everywhere. In practice I don’t think anyone can economically host the whole model plus the kv cache for the entire context size of 128k (and I’m skeptical of Deepseek’s claims now anyway). Edit: a Kagi team member just said on Discord that they’ll be increasing max tokens next release [1] https://help.kagi.com/kagi/ai/llms-privacy.html https://help.kagi.com/kagi/ai/llms-privacy.html