specific method · filed under inference & serving
Preserved thinking history mode
A history mode that retains prior reasoning content, including reasoning content in assistant messages across multi-turn conversations and tool calls.
Also called preserved thinking, Thinking Preservation.
- sources
- 7
- models
- 4
- labs adopt it
- 3
- strongest
- core
How sources treat it
One count per evidence span, weakest treatment to strongest.
Documented in
Further reading
Picked by hand, not extracted: where to read more, not evidence for anything on this page.
- Thinking Mode - DeepSeek API Docs docs api-docs.deepseek.comdocuments passing reasoning_content back across tool-call turns
Evidence
7 spans quoted from the sources, strongest treatment first.
Kimi K3 was trained in the preserved thinking history mode. For multi-turn conversations and tool calls, Kimi K3 requires the complete assistant message returned by the API to be passed back to `messages` as-is — including `reasoning_content` and `tool_calls`, not just `content`.
the think channel is always retained in the history—kept even when its content is empty—so that the model observes a consistent message structure across turns
By default, Qwen3.8-Flash-Next retains thinking blocks from all historical messages, maintaining a complete reasoning trace across the conversation.
A new feature retains thinking context across conversation history, streamlining iterative development and reducing overhead.
keep reasoning_content from prior assistant messages in the message history
K3 was trained in the preserved thinking history mode.
For multi-turn conversations and tool calls, Kimi K3 requires the complete assistant message returned by the API to be passed back to messages as-is — including reasoning_content and tool_calls, not just content
Filed alongside
Other methods under inference & serving :: context management.