mirror of
https://github.com/coder/coder.git
synced 2026-09-24 15:04:27 +08:00
docs(docs/ai-coder/ai-gateway/clients): update VS Code to reflect 1.122 Custom Endpoint support (#26126)
Updates the VS Code AI Gateway client docs to reflect the Custom Endpoint provider introduced in VS Code 1.121 (Insiders) and promoted to Stable in 1.122. ## Changes **`docs/ai-coder/ai-gateway/clients/vscode.md`** - Replace the deprecated `customoai` vendor with `customendpoint` - Add the required `apiType` field (`responses` for OpenAI, `messages` for Anthropic) - Add Anthropic provider setup (Messages API type, base URL `…/aibridge/anthropic`) - Note GitHub sign-in is no longer required — works in air-gapped/restricted environments - Add limitation callout: inline suggestions and NES still require GitHub Copilot - Reflect the UI-first API key entry flow (VS Code stores the token securely; do not paste into JSON directly) - Drop the Centralized/BYOK split — VS Code has no template injection path, so both scenarios follow the same user-driven UI flow **`docs/ai-coder/ai-gateway/clients/index.md`** - VS Code compatibility row: Anthropic `❌ → ✅` - Updated Notes column <details> <summary>Research notes</summary> - VS Code 1.121 shipped the Custom Endpoint provider (Insiders), replacing the legacy OpenAI Compatible (`customoai`) provider which is now deprecated. - VS Code 1.122 promoted Custom Endpoint to Stable and removed the GitHub sign-in requirement for BYOK. - Anthropic support confirmed working: `apiType: "messages"` + base URL `…/api/v2/aibridge/anthropic` (Coder's gateway accepts the Coder session token as `x-api-key`). - OpenAI uses `apiType: "responses"` + base URL `…/api/v2/aibridge/openai`. - API keys must be entered via the Manage Language Models UI — VS Code stores them securely and references them as `${input:chat.lm.secret.XXXXX}` in the JSON. </details> > This PR was drafted by Coder Agents on behalf of @matifali.
This commit is contained in:
Atif Ali
@@ -44,7 +44,7 @@ The table below shows tested AI clients and their compatibility with AI Gateway.
|
||||
| [Factory](./factory.md) | ✅ | ✅ | ✅ | |
|
||||
| [Cline](./cline.md) | ✅ | ✅ | ✅ | |
|
||||
| [Kilo Code](./kilo-code.md) | ✅ | ✅ | ❌ | |
|
||||
| [VS Code](./vscode.md) | ✅ | ❌ | ❌ | Only supports Custom Base URL for OpenAI. |
|
||||
| [VS Code](./vscode.md) | ✅ | ✅ | ❌ | VS Code 1.122+ via Custom Endpoint provider. GitHub sign-in not required. Inline suggestions still require GitHub Copilot. |
|
||||
| [JetBrains IDEs](./jetbrains.md) | ✅ | ❌ | ❌ | Works in Chat mode via [third-party model configuration](https://www.jetbrains.com/help/ai-assistant/use-custom-models.html#provide-your-own-api-key). |
|
||||
| [Zed](./zed.md) | ✅ | ✅ | ❌ | |
|
||||
| [GitHub Copilot](./copilot.md) | ⚙️ | - | - | Requires [AI Gateway Proxy](../ai-gateway-proxy/index.md). Uses per-user GitHub tokens. |
|
||||
|
||||
@@ -1,60 +1,77 @@
|
||||
# VS Code
|
||||
|
||||
> [!NOTE]
|
||||
> AI Gateway requires the [AI Governance Add-On](../../ai-governance.md).
|
||||
> As of Coder v2.32, deployments without the add-on will not be able to
|
||||
> access AI Gateway.
|
||||
|
||||
VS Code's native chat can be configured to use AI Gateway with the GitHub Copilot Chat extension's custom language model support.
|
||||
VS Code's native chat can be configured to use AI Gateway via the **Custom Endpoint** language model provider (VS Code 1.122+, Stable). GitHub sign-in is not required, so this works in air-gapped or restricted environments.
|
||||
|
||||
## Centralized API Key
|
||||
## Setup
|
||||
|
||||
Requires VS Code 1.122+ and the [GitHub Copilot Chat extension](https://marketplace.visualstudio.com/items?itemName=GitHub.copilot-chat).
|
||||
|
||||
For each provider below, the setup steps are:
|
||||
|
||||
1. Open the Command Palette (`Ctrl+Shift+P` / `Cmd+Shift+P` on Mac) and run **Chat: Manage Language Models**.
|
||||
1. Select **Add** → **Custom Endpoint**.
|
||||
1. Enter a **group name**, **display name**, your **[Coder API token](../../../admin/users/sessions-tokens.md#generate-a-long-lived-api-token-on-behalf-of-yourself)** as the API key, and the **API type** shown below.
|
||||
1. To add or edit models, select the gear icon next to the provider in the Language Models view to open `chatLanguageModels.json`.
|
||||
|
||||
> [!IMPORTANT]
|
||||
> You need the **Pre-release** version of the [GitHub Copilot Chat extension](https://marketplace.visualstudio.com/items?itemName=GitHub.copilot-chat) and [VS Code Insiders](https://code.visualstudio.com/insiders/).
|
||||
> Enter your API token through the UI. VS Code stores it securely and inserts a reference like `${input:chat.lm.secret.XXXXX}` into the JSON. Do not paste your token directly into the JSON file.
|
||||
|
||||
1. Open command palette (`Ctrl+Shift+P` or `Cmd+Shift+P` on Mac) and search for _Chat: Open Language Models (JSON)_.
|
||||
1. Paste the following JSON configuration, replacing `<your-coder-api-token>` with your **[Coder API token](../../../admin/users/sessions-tokens.md#generate-a-long-lived-api-token-on-behalf-of-yourself)**:
|
||||
_Replace `coder.example.com` with your Coder deployment URL. Model IDs must match what is configured in your AI Gateway._
|
||||
|
||||
### OpenAI-compatible models
|
||||
|
||||
Set **API type** to `responses`.
|
||||
|
||||
```json
|
||||
[
|
||||
{
|
||||
"name": "Coder",
|
||||
"vendor": "customoai",
|
||||
"apiKey": "<your-coder-api-token>",
|
||||
"models": [
|
||||
{
|
||||
"name": "GPT 5.2",
|
||||
"url": "https://coder.example.com/api/v2/aibridge/openai/v1/chat/completions",
|
||||
"toolCalling": true,
|
||||
"vision": true,
|
||||
"thinking": true,
|
||||
"maxInputTokens": 272000,
|
||||
"maxOutputTokens": 128000,
|
||||
"id": "gpt-5.2"
|
||||
},
|
||||
{
|
||||
"name": "GPT 5.2 Codex",
|
||||
"url": "https://coder.example.com/api/v2/aibridge/openai/v1/responses",
|
||||
"toolCalling": true,
|
||||
"vision": true,
|
||||
"thinking": true,
|
||||
"maxInputTokens": 272000,
|
||||
"maxOutputTokens": 128000,
|
||||
"id": "gpt-5.2-codex"
|
||||
}
|
||||
]
|
||||
}
|
||||
]
|
||||
{
|
||||
"name": "Coder (OpenAI)",
|
||||
"vendor": "customendpoint",
|
||||
"apiKey": "${input:chat.lm.secret.XXXXX}",
|
||||
"apiType": "responses",
|
||||
"models": [
|
||||
{
|
||||
"id": "gpt-5.5",
|
||||
"name": "GPT 5.5",
|
||||
"url": "https://coder.example.com/api/v2/aibridge/openai",
|
||||
"toolCalling": true,
|
||||
"vision": true,
|
||||
"thinking": true,
|
||||
"streaming": true,
|
||||
"maxInputTokens": 272000,
|
||||
"maxOutputTokens": 128000
|
||||
}
|
||||
]
|
||||
}
|
||||
```
|
||||
|
||||
_Replace `coder.example.com` with your Coder deployment URL._
|
||||
### Anthropic models
|
||||
|
||||
> [!NOTE]
|
||||
> The setting names may change as the feature moves from pre-release to stable. Refer to the official documentation for the latest setting keys.
|
||||
Set **API type** to `messages`.
|
||||
|
||||
## BYOK (Personal API Key)
|
||||
```json
|
||||
{
|
||||
"name": "Coder (Anthropic)",
|
||||
"vendor": "customendpoint",
|
||||
"apiKey": "${input:chat.lm.secret.XXXXX}",
|
||||
"apiType": "messages",
|
||||
"models": [
|
||||
{
|
||||
"id": "claude-sonnet-4.6",
|
||||
"name": "Claude Sonnet 4.6",
|
||||
"url": "https://coder.example.com/api/v2/aibridge/anthropic",
|
||||
"toolCalling": true,
|
||||
"vision": true,
|
||||
"thinking": true,
|
||||
"streaming": true,
|
||||
"maxInputTokens": 1000000,
|
||||
"maxOutputTokens": 64000
|
||||
}
|
||||
]
|
||||
}
|
||||
```
|
||||
|
||||
> [!NOTE]
|
||||
> At the time of writing, GitHub Copilot Chat does not support sending custom headers, so BYOK mode is not available.
|
||||
|
||||
**References:** [GitHub Copilot - Bring your own language model](https://code.visualstudio.com/docs/copilot/customization/language-models#_add-an-openaicompatible-model)
|
||||
**References:** [VS Code - Bring your own language model](https://code.visualstudio.com/docs/copilot/customization/language-models)
|
||||
|
||||
Reference in New Issue
Block a user