Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.requesty.ai/llms.txt

Use this file to discover all available pages before exploring further.

GitHub Copilot for Visual Studio Code supports Bring Your Own Key (BYOK) through a Custom Endpoint provider. By pointing that endpoint at Requesty, you can drive Copilot Chat, tools, and MCP servers with any model in the Requesty Model Library. Using the Requesty integration, you can:
  • Access 300+ models from OpenAI, Anthropic, Google, Mistral, and many other providers through one API key.
  • Track and manage your spend in a single location.
  • Apply fallback policies, load balancing, and latency routing to keep Copilot responsive.

​
Requirements

This integration requires Visual Studio Code 1.122 or later. The Custom Endpoint provider reached Stable, and BYOK now works without a GitHub sign-in, in the 1.122 release (May 28, 2026). Use Check for Updates in VS Code to upgrade.
Copilot’s inline suggestions and next edit suggestions (NES) still require a GitHub sign-in. BYOK powers chat, tools, and MCP servers only.

​
Prerequisites

  • VS Code 1.122 or later with the GitHub Copilot Chat extension installed.
  • A Requesty API key from the API Keys page.

​
Configuration

​
1. Open Manage Language Models

In the Chat view, open the model picker and select the gear icon, then Manage Language Models. You can also run Chat: Manage Language Models from the Command Palette. Manage Language Models from the model picker

​
2. Add Requesty as a Custom Endpoint

Select Add Models… and choose Custom Endpoint from the provider list. Add a Custom Endpoint provider When prompted, enter the Requesty endpoint and credentials:
FieldValue
URLhttps://router.requesty.ai
API KeyYour Requesty API key (rqsty-sk-...)

​
3. Choose the API type

VS Code asks which API the endpoint speaks: Chat Completions, Responses, or Messages. We recommend choosing messages.
Use the messages API type to get the most out of Anthropic models, including native web search. Anthropic’s native web search tool is exposed through the Messages API, so choosing Chat Completions or Responses will not give you that capability.
VS Code stores the provider in your settings.json. The result looks like this: Requesty Custom Endpoint in settings.json 
{
  "name": "requesty",
  "vendor": "customendpoint",
  "apiKey": "${input:chat.lm.secret.xxxxxxxx}",
  "apiType": "messages",
  "models": [
    {
      "id": "openai/gpt-5.5",
      "name": "Gpt-5.5",
      "url": "https://router.requesty.ai",
      "toolCalling": true,
      "vision": true,
      "maxInputTokens": 1000000,
      "maxOutputTokens": 128000
    }
  ]
}
Set each model’s id to a model from the Model Library (for example anthropic/claude-sonnet-4-6) and point url at https://router.requesty.ai. You can add multiple entries to the models array.

​
4. Select a model

After adding the provider, your Requesty models appear in the model picker tagged with Requesty. Pick one to start chatting. You can choose any model from the Model Library. Standard model IDs follow the format provider/model-name:
  • anthropic/claude-sonnet-4-6
  • openai/gpt-5.4
Policies follow the format policy/policy-name:
  • policy/reliable-coding
See Routing Policies for how to create a policy that automatically falls back between models.

​
Utility models

A few flows in VS Code, such as chat title generation, commit message generation, and feedback, use a smaller utility model that normally comes from your Copilot subscription. When you use BYOK while signed out, point those at a Requesty model too:
  1. When the notification appears in the chat input, select Configure.
  2. Set both chat.utilityModel and chat.utilitySmallModel to one of your Requesty models.
This unlocks the full set of AI features using your own key. If you only need chat, you can dismiss the notification and leave the utility-driven features inactive.

​
Verification

Send a simple prompt in Copilot Chat. The expected output is a working response from the model. Then check the Requesty analytics dashboard to confirm the request was logged.

​
Troubleshooting

The Chat view becomes available once at least one BYOK model is configured. Add the Requesty Custom Endpoint and reload the window if needed.
On VS Code 1.122+, sign-in prompts are suppressed once a BYOK model is configured. Confirm your VS Code version with Help → About and update if it is below 1.122.
Send a unique, traceable prompt and re-check the analytics dashboard. Confirm the URL is https://router.requesty.ai and the API key is valid.
If your Requesty models are missing from the model picker, reopen the Manage Language Models view from the Command Palette:
  1. Open the Command Palette: Cmd+Shift+P (Mac) or Ctrl+Shift+P (Windows/Linux).
  2. Type: Chat: Manage Language Models and select it.
Re-add or re-select the Requesty Custom Endpoint, then reload the window if the models still don’t show.

​
Security

Do not commit your API key into shared docs or version-controlled files. VS Code stores BYOK credentials locally.

​
References

Last modified on May 31, 2026