Unhandled API in mapOptionsForApi: undefined at mapOptionsForApi
CairbreMadog
OP
 — 2026-01-27, 17:13
I am trying to point a fresh install of clawdbot at ollama server on my network. I have tried the GLM 4.7 Flash model I have seen tutorials for, but I cant seem to point the config at the right place to not crash the gateway server with "Unhandled API in mapOptionsForApi: undefined  at mapOptionsForApi"


Here is an example:= of my JSON w/signal configured on my other signal-cli vm:

 {
  "channels": {
    "signal": {
      "enabled": true,
      "account": "+1XXXXXXXXXX",
      "httpUrl": "http://192.168.1.xx:8090/",
      "autoStart": false,
      "allowFrom": ["+1XXXXXXXXXX"]
    }
  },
  "models": {
    "providers": {
      "ollama": {
        "baseUrl": "http://192.168.2.xx:11434/v1",
        "apiKey": "ollama-local",
        "api": "openai-completions",
        "models": [
          {
            "id": "glm-4.7-flash",
            "name": "GLM 4.7 Flash",
            "reasoning": false,
            "input": ["text"],
            "contextWindow": 108192,
            "maxTokens": 81920,
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 }
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": { "primary": "ollama/glm-4.7-flash" }
    }
  }
}
Answer Overflow
APP
 — 2026-01-27, 17:13
To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution
Image
Krill 🦐:lobster:
APP
 — 2026-01-27, 17:15
Found it! The error Unhandled API in mapOptionsForApi: undefined means the model is missing its api field. You're defining api: "openai-completions" at the provider level, but each model needs its own api field.

Fix your model definition:

{
  "models": {
    "providers": {
      "ollama": {
        "baseUrl": "http://192.168.2.XX:11434/v1",
        "apiKey": "ollama-local",
        "models": [
          {
            "id": "glm-4.7-flash",
            "name": "GLM 4.7 Flash",
            "api": "openai-completions",  // ← ADD THIS to each model
            "reasoning": false,
            "input": ["text"],
            "contextWindow": 108192,
            "maxTokens": 81920,
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 }
          }
        ]
      }
    }
  }
}


The provider-level api field isn't used — only the model-level one matters for routing to mapOptionsForApi().

📚 Docs: https://docs.clawd.bot/providers/ollama#explicit-setup-manual-models
CairbreMadog
OP
 — 2026-01-27, 17:53
cool no longer getting the error.. it give me these messages now when I send anything from the gateway: 22:52:20 [agent/embedded] embedded run done: runId=a9cc829b-c5bb-46ac-8808-144ebedcc930 sessionId=80b24506-1c2b-436d-9749-b28885df997b durationMs=32 aborted=false
22:52:20 [diagnostic] lane task done: lane=main durationMs=33 active=0 queued=0
22:52:20 [diagnostic] lane task done: lane=session:agent:main:main durationMs=34 active=0 queued=0

In the gateway I get no response and my ollama server doesn't react. 
Krill 🦐:lobster:
APP
 — 2026-01-27, 17:54
The 32ms completion suggests the model call isn't happening. A few things to check:

Verify the model name exists on your Ollama server:


# On the Ollama host (192.168.2.XX)
ollama list


GLM models aren't standard Ollama models. You likely need to either:
Use a proper Ollama model like llama3.3, qwen2.5-coder:32b, etc.
Or if you're trying to use a remote GLM API, that should be a different provider (OpenAI-compatible endpoint), not Ollama

Test connectivity to Ollama:


bash
curl http://192.168.2.XX:11434/api/tags


If you meant to use a remote GLM API (not local Ollama):
 you'd use the OpenAI provider pointing to the GLM endpoint, not the Ollama provider.

{
  "models": {
    "providers": {
      "openai-compatible": {  // ← different provider
        "baseUrl": "https://open.bigmodel.cn/api/paas/v4",  // GLM endpoint
        "apiKey": "YOUR_API_KEY",
        "models": [
          {
            "id": "glm-4.7-flash",
            "api": "openai-completions",
            // ...
          }
        ]
      }
    }
  }
}


What's the output of ollama list on your 192.168.2.XX machine?
CairbreMadog
OP
 — 2026-01-27, 17:55
NAME                           ID              SIZE      MODIFIED
openchat:latest                537a4e03b649    4.1 GB    7 hours ago
glm-4.7-flash:latest           d1a8a26252f1    19 GB     7 hours ago
llama3.2:latest                a80c4f17acd5    2.0 GB    6 weeks ago
llama3:latest                  365c0bd3c000    4.7 GB    6 weeks ago
dominus-maximus:latest         374bab6c03d0    5.0 GB    2 months ago
gemma:7b                       a72c7f4d0a15    5.0 GB    2 months ago
bge-m3:latest                  790764642607    1.2 GB    3 months ago
llama3.1:8b-instruct-q5_K_M    27fe1b0ab52c    5.7 GB    3 months ago
CairbreMadog
OP
 — 2026-01-28, 08:43
@Krill 🦐  Can you direct me to a good place that has excellent documentation on setting up a remote ollama instance. It just doesn't work for me