[GH-ISSUE #599] OAI Compatible Provider for Copilot集成文档应该根据V4 API更新 #172

New issue

Open

opened 2026-10-26 09:54:43 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 09:54:43 +00:00

Owner

Originally created by @minecx on GitHub (Apr 29, 2026).
Original GitHub issue: https://github.com/deepseek-ai/awesome-deepseek-integration/issues/599

原文件最后一段现在显示的还是DeepSeek上一版模型的id和上下文长度：

"oaicopilot.baseUrl": "https://api.deepseek.com/v1",
"oaicopilot.models": [
  {
      "id": "deepseek-chat",
      "owned_by": "deepseek",
      "context_length": 128000,
      "max_tokens": 8000
  },
  {
      "id": "deepseek-reasoner",
      "owned_by": "deepseek",
      "context_length": 128000,
      "max_tokens": 16000
  }
]

建议改为：

"oaicopilot.models": [
    {
        "id": "deepseek-v4-flash",
        "owned_by": "deepseek",
        "configId": "DS V4 Flash (High)",
        "displayName": "DS V4 Flash (High)",
        "context_length": 1000000,
        "max_tokens": 384000,
        "include_reasoning_in_request": true,
        "thinking": {
            "type": "enabled"
        },
        "reasoning_effort": "high"
    },
    {
        "id": "deepseek-v4-flash",
        "owned_by": "deepseek",
        "configId": "DS V4 Flash (Max)",
        "displayName": "DS V4 Flash (Max)",
        "context_length": 1000000,
        "max_tokens": 384000,
        "include_reasoning_in_request": true,
        "thinking": {
            "type": "enabled"
        },
        "reasoning_effort": "max"
    },
    {
        "id": "deepseek-v4-pro",
        "owned_by": "deepseek",
        "configId": "DS V4 Pro (High)",
        "displayName": "DS V4 Pro (High)",
        "context_length": 1000000,
        "max_tokens": 384000,
        "include_reasoning_in_request": true,
        "thinking": {
            "type": "enabled"
        },
        "reasoning_effort": "high"
    },
    {
        "id": "deepseek-v4-pro",
        "owned_by": "deepseek",
        "configId": "DS V4 Pro (Max)",
        "displayName": "DS V4 Pro (Max)",
        "context_length": 1000000,
        "max_tokens": 384000,
        "include_reasoning_in_request": true,
        "thinking": {
            "type": "enabled"
        },
        "reasoning_effort": "max"
    }
],

注意：

在使用此API时需要加上"include_reasoning_in_request": true，否则插件不会将上一轮的reasoning_content自动拼接到下一轮的请求中，从而导致DeepSeek API报错。
模型中的id必须和DeepSeek官方API中的id相符，否则会提示找不到模型。但为了区分pro / flash模型和思考强度，我自己加入了configId, displayName, thinking & reasoning_effort。其余用户可以根据自己的需要修改或删减相关参数。

Originally created by @minecx on GitHub (Apr 29, 2026). Original GitHub issue: https://github.com/deepseek-ai/awesome-deepseek-integration/issues/599 [原文件](https://github.com/deepseek-ai/awesome-deepseek-integration/blob/main/docs/oai-compatible-copilot/README_cn.md)最后一段现在显示的还是DeepSeek上一版模型的id和上下文长度： ``` "oaicopilot.baseUrl": "https://api.deepseek.com/v1", "oaicopilot.models": [ { "id": "deepseek-chat", "owned_by": "deepseek", "context_length": 128000, "max_tokens": 8000 }, { "id": "deepseek-reasoner", "owned_by": "deepseek", "context_length": 128000, "max_tokens": 16000 } ] ``` 建议改为： ``` "oaicopilot.models": [ { "id": "deepseek-v4-flash", "owned_by": "deepseek", "configId": "DS V4 Flash (High)", "displayName": "DS V4 Flash (High)", "context_length": 1000000, "max_tokens": 384000, "include_reasoning_in_request": true, "thinking": { "type": "enabled" }, "reasoning_effort": "high" }, { "id": "deepseek-v4-flash", "owned_by": "deepseek", "configId": "DS V4 Flash (Max)", "displayName": "DS V4 Flash (Max)", "context_length": 1000000, "max_tokens": 384000, "include_reasoning_in_request": true, "thinking": { "type": "enabled" }, "reasoning_effort": "max" }, { "id": "deepseek-v4-pro", "owned_by": "deepseek", "configId": "DS V4 Pro (High)", "displayName": "DS V4 Pro (High)", "context_length": 1000000, "max_tokens": 384000, "include_reasoning_in_request": true, "thinking": { "type": "enabled" }, "reasoning_effort": "high" }, { "id": "deepseek-v4-pro", "owned_by": "deepseek", "configId": "DS V4 Pro (Max)", "displayName": "DS V4 Pro (Max)", "context_length": 1000000, "max_tokens": 384000, "include_reasoning_in_request": true, "thinking": { "type": "enabled" }, "reasoning_effort": "max" } ], ``` 注意： 1. 在使用此API时需要加上`"include_reasoning_in_request": true`，否则插件不会将上一轮的`reasoning_content`自动拼接到下一轮的请求中，从而导致DeepSeek API报错。 2. 模型中的id必须和DeepSeek官方API中的id相符，否则会提示找不到模型。但为了区分pro / flash模型和思考强度，我自己加入了`configId`, `displayName`, `thinking` & `reasoning_effort`。其余用户可以根据自己的需要修改或删减相关参数。

bsakaguchi referenced this issue

2026-10-26 09:56:41 +00:00

[PR #172] [CLOSED] feat: fix markdown rendering #326