HTTP 400 Error Code 1211 Unknown Model - GLM 5.2

[fossfann]

Hi there,

When i configured attacker brain with z.ai coding subscription API key i am getting the shared error. When i checked my api key with curl command i am getting successful response from z.ai like below:

curl -sS 'https://api.z.ai/api/coding/paas/v4/chat/completions'
-H "Authorization: Bearer YOURAPIKEY"
-H 'Content-Type: application/json'
-d '{"model":"glm-5.2","messages":[{"role":"user","content":"Only say hello world"}],"max_tokens":64,"thinking":{"type":"disabled"}}'
{"choices":[{"finish_reason":"stop","index":0,"message":{"content":"Hello world","role":"assistant"}}],"created":1784567414,"id":"2026072101101193c6a2969f3c4b65","model":"glm-5.2","object":"chat.completion","request_id":"2026072101101193c6a2969f3c4b65","usage":{"completion_tokens":3,"completion_tokens_details":{"reasoning_tokens":0},"prompt_tokens":10,"prompt_tokens_details":{"cached_tokens":0},"total_tokens":13}}

EDIT: I wrote GLMMM-5.2 on purpose and it replied as "unknown model". I think this means API key, base URL and model name is correct.

curl -sS 'https://api.z.ai/api/coding/paas/v4/chat/completions' -H "Authorization: Bearer YOURAPIKEY" -H 'Content-Type: application/json' -d '{"model":"GLMMM-5.2","messages":[{"role":"user","content":"Only say hello world"}],"max_tokens":64,"thinking":{"type":"disabled"}}'
{"error":{"code":"1211","message":"Unknown Model, please check the model code."}}

i checked from z.ai coding subscription API key documentation and base URL is https://api.z.ai/api/coding/paas/v4 and my attacker brain config is like below:

protocol = "openai"
base_url = "https://api.z.ai/api/coding/paas/v4"
api_key = "........."
model = "glm-5.2"

[fossfann]

i solved the issue. here the fix.

Open wallbreaker/providers/openai_provider.py and find this exact block below:

    base_url = getattr(self.endpoint, "base_url", "") or ""
    if getattr(self.endpoint, "cache", True) and "openrouter.ai" in base_url:
        _apply_openrouter_cache(
            wire_messages, getattr(self.endpoint, "cache_ttl", "5m")
        )
    payload: dict = {
        "model": self.endpoint.model,
        "messages": wire_messages,
        "max_tokens": max_tokens,
        "stream": True,
    }

and replace with below:

    base_url = getattr(self.endpoint, "base_url", "") or ""
    if getattr(self.endpoint, "cache", True) and "openrouter.ai" in base_url:
        _apply_openrouter_cache(
            wire_messages, getattr(self.endpoint, "cache_ttl", "5m")
        )
    # Strip OpenRouter provider prefix for Z.AI direct API
    _model = self.endpoint.model
    if "api.z.ai" in base_url and "/" in _model:
        _model = _model.split("/", 1)[1]
    
    payload: dict = {
        "model": _model,
        "messages": wire_messages,
        "max_tokens": max_tokens,
        "stream": True,
    }

this fix is stripping z-ai/ and HTTP requests are going NOT "z-ai/glm-5.2" and going as "glm-5.2"

by the way in the TUI i can see z-ai/glm-5.2 in the right section but this can be fix by modifying some file.