Deepseek的max tokens貌似无法限制长度。

[denggit]

配置如下：
{
"name": "deepseek",
"api_base_url": "https://api.deepseek.com/chat/completions",
"api_key": "*",
"models": ["deepseek-chat", "deepseek-reasoner"],
"transformer": {
"use": [
"deepseek",
["maxtoken", { "max_tokens": 123000 }]
],
"deepseek-chat": {
"use": ["tooluse"],
"completion_tokens": 8000
}
}
},

错误如下：
API Error: 400 {"error":{"message":"Error from provider(deepseek,deepseek-chat: 400): {"error":{"message":"This model's maximum context length is 131072 tokens. However, you
requested 131329 tokens (123137 in the messages, 8192 in the completion). Please reduce the length of the messages or
completion.","type":"invalid_request_error","param":null,"code":"invalid_request_error"}}Error: Error from provider(deepseek,deepseek-chat: 400):
{"error":{"message":"This model's maximum context length is 131072 tokens. However, you requested 131329 tokens (123137 in the messages, 8192 in the completion). Please reduce
the length of the messages or completion.","type":"invalid_request_error","param":null,"code":"invalid_request_error"}}\n at nt
(/opt/homebrew/lib/node_modules/@musistudio/claude-code-router/dist/cli.js:75644:11)\n at u0 (/opt/homebrew/lib/node_modules/@musistudio/claude-code-router/dist/cli.js:75702:11)\n
at process.processTicksAndRejections (node:internal/process/task_queues:105:5)\n at async o0
(/opt/homebrew/lib/node_modules/@musistudio/claude-code-router/dist/cli.js:75669:84)","type":"api_error","code":"provider_response_error"}}

[starship863]

我同样有类似困扰，区别是我用的是qwen。千问不会报错，但是一晚上消耗了我800元。具体配置文件如下：

{
"APIKEY": "sk-xxx",
"PROXY_URL": "http://127.0.0.1:7890",
"LOG": true,
"API_TIMEOUT_MS": 600000,
"NON_INTERACTIVE_MODE": false,

"Providers": [
{
"name": "deepseek",
"api_base_url": "https://api.deepseek.com/chat/completions",
"api_key": "sk-xxxx",
"models": ["deepseek-chat", "deepseek-reasoner"],
"transformer": {
"use": ["deepseek"],
"deepseek-chat": {
"use": ["tooluse"]
}
}
},
{
"name": "dashscope",
"api_base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions",
"api_key": "sk-xxxx",
"models": ["qwen3-coder-plus"],
"transformer": {
"use": [
[
"maxtoken",
{
"max_tokens": 65536
}
],
"enhancetool"
]
}
}
],

"Router": {
"default": "dashscope,qwen3-coder-plus", // 默认模型
"background": "dashscope,qwen3-coder-plus", // 后台任务
"think": "dashscope,qwen3-coder-plus", // 推理/思考流程
"longContext": "dashscope,qwen3-coder-plus", // 长文本处理
"longContextThreshold": 60000, // 触发长上下文的 token 阈值
"webSearch": "dashscope,qwen3-coder-plus" // Web 搜索后处理
}
}