Skip to content
Source
Sign inSign up
TanStack/schemas

Fill replayReasoningContent for DeepSeek, Kimi, GLM, and Qwen hosts #270

Send to my agent
Closedtombeckenham opened this issue 1 comment
tombeckenhamcommented

Problem

requestMap.replayReasoningContent is true on only 20 rows (DeepSeek and Fireworks). These models also need reasoning_content on earlier assistant turns, or the second turn fails. Measured 2026-10-09; models.dev interleaved.field: "reasoning_content" used as a hint:

ProviderRowsExamples
baseten11deepseek-ai/DeepSeek-V4-Pro-0813, moonshotai/Kimi-K3, zai-org/GLM-5.2
huggingface11deepseek-ai/DeepSeek-V4-Pro, MiniMaxAI/MiniMax-M2.7, moonshotai/Kimi-K2.6
dashscope9qwen3.8-max, kimi-k3, deepseek-v4-flash-0731
openrouter7deepseek/deepseek-v4-pro, z-ai/glm-5.2
together7deepseek-ai/DeepSeek-V4.1-Flash, zai-org/GLM-5.2
zai7glm-5.2, glm-5.3-flash, glm-4.7
nvidia6z-ai/glm-5.3, moonshotai/kimi-k3
moonshotai-cn4kimi-k2.6, kimi-k3
moonshot2kimi-k2.6, kimi-k2.7-code
cloudflare-workers-ai2@cf/moonshotai/kimi-k2.6
zhipuai-coding-plan2glm-5.3
cloudflare-ai-gateway1alibaba/qwen3.8-max

Ask

Fill replayReasoningContent from each provider's docs. The model makers document it (DeepSeek, Moonshot, Z.AI thinking docs). A host that serves the same model on Chat Completions usually needs the same thing; confirm it from the host's docs before setting it.

Acceptance

  • Every row in the table has replayReasoningContent set to true or false, or a ledger entry.
  • moonshot/kimi-k2.6 and zai/glm-5.2 are true.
  1. tombeckenham added labelsbug

  2. Completed through reviewed provider-native extraction and PR #293. Production verification after the scheduled poll finds zero unaccounted replayReasoningContent gaps across the issue's original 12 providers. Both required rows are sourced true: moonshot/kimi-k2.6 and zai/glm-5.2.

    Replay booleans are emitted only from the serving provider's published contract or operational guidance. Where an exact model's healthy native source omits a replay rule, its value stays null with scoped checked-source evidence. Provider-wide guessed recipes were removed. NVIDIA's refreshed contracts preserve exact selectors and published endpoint paths; malformed/unavailable sources remain errors.

    Recorded source URLs, hashes, and model-scoped omission evidence are now visible in the public API by default and on the web pages. Upstream captures, all ledger scopes, Worker ingestion/repoll tests, independent exact-head reviews, all 1,881 tests, and required CI passed.

  3. tombeckenham closed this

Sign in to join this conversation.