Unverified70% confidenceFactExact time
The core change of this update is to retain reasoning content across multi-turn requests by default on supported models.
2
Sources
70%
Confidence
Medium-term (~90 days)
Relevance
9/17/2026
First Seen
Valid until: 12/16/2026
Sources
Related Entities
Related Claims
UnverifiedThis update supports specifying different models for each individual request within a batch request (per-request models).76% similarUnverifiedThe version under test is jointly determined by the model used, model parameters, prompts, and context; a change in any one of these constitutes a change in the version under test.71% similarUnverifiedUpdate merging refers to independently computing, for each target language, the parameter update relative to the source model and then merging them into the source model all at once or in batches, with the updates for each language not interfering with one another.71% similarUnverifiedParameterized policy updates directly modify model weights (e.g., online updating of LoRA fine-tuning layers), offering a higher capability ceiling but introducing the risk of catastrophic forgetting; non-parameterized approaches change behavior by modifying prompts, retrieved content, or tool configurations, offering strong reversibility but limited expressiveness.70% similarUnverifiedMaintaining two flags, supportsReasoning and defaultsToReasoning, for each model can help the gateway correctly decide whether to inject reasoning parameters.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/928186API
curl https://kongchang.com/api/v1/knowledge/claims/928186MCP
get_claim(id=928186)