Issues

• [Bug] GLM-5.3-Flash: complete output collapse into multilingual token garbage (severe escalation of #142)

  zai-org/GLM-5#163 · Filter by author ZinkDifferent opened on Sep 22, 2026
  2

  

• [Bug] GLM-5.3-Flash chat template crashes on OpenAI-spec string tool_calls[].function.arguments (chat_template.jinja line 163)

  zai-org/GLM-5#158 · Filter by author fallow5 opened on Sep 20, 2026
  5

  

• Enhancement: Add API usage dashboard with token, cache, and cost breakdown

  enhancement
  zai-org/GLM-5#154 · Filter by author Ahad690 opened on Sep 15, 2026
  

• 在智能体工具使用的时候有问题

  zai-org/GLM-5#153 · Filter by author ydsgangge-ux opened on Sep 15, 2026
  3

  

• [Coding Plan] MCP tools (web_search_prime / web_reader) hit 429 rate limit aggressively — thresholds undocumented, unrelated to quota

  zai-org/GLM-5#151 · Filter by author Enominera opened on Sep 12, 2026
  1

  

• GLM API Usage Bundle purchased and confirmed by payment receipt, but account balance shows 0.00 — error 1113 on all requests

  zai-org/GLM-5#148 · Filter by author smabbutt opened on Sep 11, 2026
  

• [Bug] glm-5.3-flash: sending tools adds a flat ~3s to first token, per request — independent of tool count, tool_choice, and a 100% prompt-cache hit

  zai-org/GLM-5#144 · Filter by author Ahad690 opened on Sep 5, 2026
  2

  

• [Bug] GLM-5.3-Flash spontaneously emits Chinese text mid-response in English agentic contexts (survives context reset)

  zai-org/GLM-5#142 · Filter by author ZinkDifferent opened on Sep 4, 2026
  7

  

• Anthropic-compatible API: message_start reports zero usage.input_tokens, breaking clients that read context usage at stream start

  zai-org/GLM-5#139 · Filter by author nontravis opened on Sep 1, 2026
  

• GLM-5.3-Flash quotes a system prompt it was never given on video input, and sometimes answers it instead of the user

  zai-org/GLM-5#134 · Filter by author pudepiedj opened on Aug 28, 2026

• [Bug] GLM-5.2: command prefix planned in thinking is dropped from the emitted tool call — unrecoverable repetition loop (10 consecutive failures)

  zai-org/GLM-5#116 · Filter by author carrotRakko opened on Jul 21, 2026
  4

  

• [GLM-5.2][Safety] Fixed correction-insulin dose suggested when correction factor is unknown

  zai-org/GLM-5#111 · Filter by author goktugozkanmd opened on Jul 12, 2026
  2

  

• [GLM-5.2] Tool results intermittently rendered as image content blocks instead of text (context-length correlated)

  zai-org/GLM-5#110 · Filter by author AlexThor-Omega0 opened on Jul 11, 2026
  

• Remote OAuth MCP servers fail with invalid_token when defined in config.json (no OAuth provider attached to config-file transport)

  zai-org/GLM-5#106 · Filter by author spokvulcan opened on Jul 8, 2026
  1

  

• Official Support Channels for GLM API, ZCode, and Related Products / GLM API、ZCode 等产品官方反馈渠道

  bug
  zai-org/GLM-5#93 · Filter by author zRzRzRzRzRzRzR opened on Jun 27, 2026
  7

• reasoning_content 跨轮对模型不可见，多轮对话中推理链断裂

  zai-org/GLM-5#92 · Filter by author miao0044 opened on Jun 27, 2026
  2

  

• [Bug] Tool call errors with malformed argument keys: Model tried to call unavailable tool

  zai-org/GLM-5#84 · Filter by author aigehub opened on Jun 17, 2026
  

• [需求] 建议智谱官方 API 支持 OpenAI 兼容的 /responses 接口，以提升 Codex 兼容性

  enhancementhelp wanted
  zai-org/GLM-5#39 · Filter by author SimZhou opened on Mar 30, 2026
  43

  

• [Bug] Malformed tool-call JSON via NVIDIA NIM (GLM-5) causes parser failures in OpenCode

  zai-org/GLM-5#15 · Filter by author chindris-mihai-alexandru opened on Feb 16, 2026
  8