ollama 本地部署 代理连接问题

[PIG-990]

局域网访问都正常
curl -X POST http://192.168.9.111:11434/v1/chat/completions -H "Content-Type: application/json" -d "{"model":"gpt-oss:20b","messages":[{"role":"user","content":"你好"}],"stream":false}"
{"id":"chatcmpl-24","object":"chat.completion","created":1772004257,"model":"gpt-oss:20b","system_fingerprint":"fp_ollama","choices":[{"index":0,"message":{"role":"assistant","content":"您好！有什么可以帮您的吗？如果你有任何问题或想了解的主题，随时告诉我。😊","reasoning":"The user simply says "你好" (Hello). We should respond in Chinese, likely greet them back, possibly ask how they are or what they need. Should keep it friendly."},"finish_reason":"stop"}],"usage":{"prompt_tokens":68,"completion_tokens":71,"total_tokens":139}}

CLAWX 连接不上ollama
⚠️ Agent failed before reply: No API key found for provider "anthropic". Auth store: C:\Users\Laptop.openclaw\agents\main\agent\auth-profiles.json (agentDir: C:\Users\Laptop.openclaw\agents\main\agent). Configure auth for this agent (openclaw agents add ) or copy auth-profiles.json from the main agentDir. Logs: openclaw logs --follow

⚠️ Agent failed before reply: No API key found for provider "ollama". Auth store: C:\Users\Laptop.openclaw\agents\main\agent\auth-profiles.json (agentDir: C:\Users\Laptop.openclaw\agents\main\agent). Configure auth for this agent (openclaw agents add ) or copy auth-profiles.json from the main agentDir. Logs: openclaw logs --follow

[Steva0]

Ollama is working correctly on the LAN. I can successfully call it with:

curl -X POST http://192.168.9.111:11434/v1/chat/completions ...

and it returns a valid response from gpt-oss:20b.

The issue is not with Ollama itself, but with CLAWX/OpenClaw authentication. The agent fails before sending any request because it cannot find an API key for the configured providers:

• First it looks for anthropic
• Then it falls back to ollama
• Both fail due to missing entries in auth-profiles.json

It seems OpenClaw requires a provider configuration entry (even for local Ollama), otherwise it refuses to start the agent.

Could you clarify the correct auth-profiles.json structure for using a local Ollama endpoint (e.g. http://192.168.9.111:11434/)?
Is a dummy API key required?

[PIG-990]

ollama的 api key = ollama 就能正常调用了，要给一个默认的key

…

________________________________ 发件人: Steva ***@***.***> 发送时间: 2026年2月25日 17:16 收件人: ValueCell-ai/ClawX ***@***.***> 抄送: PIG-990 ***@***.***>; Author ***@***.***> 主题: Re: [ValueCell-ai/ClawX] ollama 本地部署 代理连接问题 (Discussion #165) Ollama is working correctly on the LAN. I can successfully call it with: curl -X POST http://192.168.9.111:11434/v1/chat/completions ... and it returns a valid response from gpt-oss:20b. The issue is not with Ollama itself, but with CLAWX/OpenClaw authentication. The agent fails before sending any request because it cannot find an API key for the configured providers: * First it looks for anthropic * Then it falls back to ollama * Both fail due to missing entries in auth-profiles.json It seems OpenClaw requires a provider configuration entry (even for local Ollama), otherwise it refuses to start the agent. Could you clarify the correct auth-profiles.json structure for using a local Ollama endpoint (e.g. http://192.168.9.111:11434/)? Is a dummy API key required? ― Reply to this email directly, view it on GitHub<#165 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/BFSCWRTLIBR72KCWR6TR6KL4NVR7JAVCNFSM6AAAAACV7ACDQSVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTKOJSGEYDONQ>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[PIG-990]

基础URL：http://192.168.9.111:11434/v1
模型ID：qwen3:30-a3b
apikey：ollama
这样就没问题了