Skip to content

Bug: invalid request "input item is missing type" on zai-org/GLM-5.3 #214

Closed

Description

@ThomasByr

Safe diagnostics

plugin: 0.8.2                                                                                                                                                                                                                                                                            
 host: pi 1.5.2                                                                                                                                                                                                                                                                           
 catalog CLI: 1.88.5                                                                                                                                                                                                                                                                      
 model: zai-org/GLM-5.3                                                                                                                                                                                                                                                                   
 route: unknown/openai-responses                                                                                                                                                                                                                                                          
 capabilities: bundled snapshot; host overrides may apply                                                                                                                                                                                                                                 
 price: known (verified 2026-11-11; estimate, not billing)                                                                                                                                                                                                                                
 last request: none                                                                                                                                                                                                                                                                       
 Timing measures received body chunks, not tokens. No prompts, tool content, credentials, account data or configured local paths are collected. Model IDs come from the host; review custom names before sharing.                                                                         
 model endpoint: openai-responses (auto; catalog)

Plan

Max

Minimal reproduction

first message: "hey" with model zai-org/GLM-5.3

Error and timing

Error: commandcode API error (400): {"message":"{\"error\":{\"code\":\"invalid_request\",\"details\":{},\"message\":\"input item is missing                                                                                                                                              
 type\",\"param\":null,\"request_id\":\"req_bxxx\",\"type\":\"invalid_request_error\"}}","type":"invalid_request_error"} 

Exact same setup while using deepseek/deepseek-v4.1-flash just works.
Edit Selecting chat-completions for zai-org/GLM-5.3 also works: Pi does not report any suspicious cache misses for that model on completions but I'm unsure if caching works on completions for that model.

Activity

  1. pierreraby commented on Nov 13, 2026

    @pierreraby
    Collaborator

    Thanks for the precise report — it reproduced exactly, and it turned out to be two stacked causes, not one. I probed the live Provider API with a minimal text-only request ("hey", max_output_tokens: 16) against zai-org/GLM-5.3 in four configurations:

    Request body Result
    input items as pi sends them today, no reasoning field 400 invalid_request: input item is missing type
    input items with type: "message", no reasoning field 200, full SSE stream
    input items with type: "message" + reasoning: {effort: "none"} 400 unsupported output_config.effort: none
    input items as pi sends them today + reasoning: {effort: "none"} 400 input item is missing type

    So the missing item type is real, and it is independent of reasoning — removing reasoning alone still fails (row 1). But fixing it alone is not enough either: once the input is accepted, effort: "none" is refused (row 3).

    Why the items are untagged

    The Responses body is built by the host encoder, not by this extension: pi-ai only tags non-message items (type: "message" for assistant text, function_call, function_call_output, reasoning). The leading system/developer item and every user item go out as a bare {role, content} pair. OpenAI accepts that shorthand; the adapter in front of GLM-5.3 does not. I confirmed it by driving the installed pi-ai encoder directly and dumping the body it produces for this model — the very first turn emits a single untagged {role: "user", ...} item.

    Why effort: "none" is sent

    When thinking is off, pi-ai reads the level from thinkingLevelMap and falls back to "none". Declaring off: null does not fix it: pi then clamps to the lowest allowed level and sends effort: "low", which silently forces reasoning on instead of disabling it. Omitting the reasoning field is what actually works.

    Not specific to GLM-5.3

    I swept all 74 models that advertise /responses with the same minimal request:

    • untagged input items refused by exactly one model: zai-org/GLM-5.3. Every other /responses model accepts the shorthand, so this is an adapter quirk, not a general defect of the Responses route.
    • effort: "none" refused by two models: zai-org/GLM-5.3 (unsupported output_config.effort: none) and mistral/mistral-large-4 (reasoning_effort none is not supported for this model, supported values: [high, low, ...]). mistral-large-4 accepts the shorthand fine and answers 200 with effort: "low" or with no reasoning field at all — it is broken the same way for anyone running it with thinking off, unrelated to this issue.

    Two unrelated observations from the sweep, worth separate issues: Qwen/Qwen3.6-Max-Preview and stepfun/Step-3.5-Flash are still advertised by /provider/v1/models but answer 404 Model '...' not found.

    Workaround available today

    /commandcode-endpoint zai-org/GLM-5.3 chat-completions (you already found it) — it works because Chat Completions has a different encoder. Same command fixes mistral/mistral-large-4.

    Fix

    A per-model repair table in the Responses fetch wrapper: add type: "message" to role-carrying input items for GLM-5.3, and drop a reasoning block that only says effort: "none" for both models. Everything else passes through byte-identical — other models, malformed JSON, non-object bodies, and already-tagged items are untouched. Both repairs are keyed off live evidence, so no other backend's request changes.

    Locally: tsc --noEmit, Prettier, 13 new unit tests over the rewrite, and the full npm test suite against real hosts are green. I'll open a PR shortly and link it here. Nothing has been released for this yet.

  2. pierreraby commented on Nov 13, 2026

    @pierreraby
    Collaborator

    Confirmed and fixed in #215 (merged as 6094904), which is on main now.

    Post-merge verification with the bodies the fix actually produces:

    model / body result
    zai-org/GLM-5.3, tagged items, no reasoning 200, real answer, 1 reasoning token
    mistral/mistral-large-4, untouched items, no reasoning 200, real answer, 228 reasoning tokens / 265 output

    One caveat worth knowing when you re-test by hand: mistral-large-4 reasons by default even with no reasoning field, so a small max_output_tokens gets spent entirely on reasoning and the answer arrives empty. That is not a wire failure — it shows up as 200 with 0 text deltas and a large reasoning_tokens. With a realistic budget it answers normally.

    The merged change adds type: "message" to the system/user items for GLM-5.3 and drops the inert reasoning block for both GLM-5.3 and mistral/mistral-large-4; every other model's request body is untouched. Until a release is cut, /commandcode-endpoint zai-org/GLM-5.3 chat-completions remains the workaround on 0.8.2.

    Closing this issue — the pricing gate is green again on main and a patch release will follow.

  3. added a commit that references this issue on Nov 13, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions