Requesty
GitHub Copilot

GitHub Copilot Chat on any model, with your own key

VS Code supports bring your own key through a Custom Endpoint. Point it at Requesty and Copilot Chat, tools and MCP servers run on any of 600+ models, billed per token to your Requesty balance.

Read the setup guide
VS Code 1.122 or laterMessages API typeno GitHub sign-in for chat
chatLanguageModels.json
custom endpoint
[
  {
    "name": "Requesty",
    "vendor": "customendpoint",
    "apiKey": "${input:chat.lm.secret.xxxxxxxx}",
    "apiType": "messages",
    "models": [
      {
        "id": "anthropic/claude-sonnet-4-6",
        "name": "Claude Sonnet 4.6",
        "url": "https://router.requesty.ai",
        "toolCalling": true
      }
    ]
  }
]
Setup

Set up GitHub Copilot in three steps

Since VS Code 1.122, bring your own key works without a GitHub sign-in for chat, tools and MCP servers. Inline suggestions and next edit suggestions still need one.

  1. 01
    Open Manage Language Models

    In the Chat view, open the model picker, select the gear icon, then Manage Language Models.

  2. 02
    Add a Custom Endpoint

    Choose Add Models, then Custom Endpoint. Enter Requesty as the group name, your Requesty API key, and Messages as the API type.

  3. 03
    Add your models

    In chatLanguageModels.json, set each id to a Requesty model ID and url to https://router.requesty.ai, without /v1. The models appear in the picker tagged Requesty.

GitHub Copilot setup in the docsCreate an API key
Models

Put any model in the Copilot picker

Each entry in chatLanguageModels.json is one model in the picker. Use a provider/model ID from the Model Library, or a policy/ name for automatic fallback between models.

The Messages API type carries native Anthropic features that Chat Completions does not.

Browse the Model LibraryRouting policies
Copilot model picker
via requesty
Model IDs that work here
anthropic/claude-sonnet-4-6Anthropic
anthropic/claude-opus-5-5Anthropic
openai/gpt-5.4OpenAI
openai/gpt-6.1-solOpenAI
policy/reliable-codingyour own routing policy
99.25%

eventual success on a routing policy in April 2026, against 85.01% for callers pinned to one provider.

Open data note

65.8%

of provider failures on the gateway in April 2026 were 429s. Throttling, not downtime, is what stops an agent mid-task.

Open data note

600+

models you can add to the Copilot picker with one Requesty key, from every major provider.

Spend controls

Copilot spend you can see and cap

Requests from Copilot Chat land in the Requesty analytics dashboard with model, tokens and cost. Give each developer a key with a monthly cap, and use a policy so a throttled provider does not stall the chat.

  • A key per developerCap it, watch it, revoke it on its own
  • Fallback policiesUse policy/ names as model IDs
  • Key kept out of the fileVS Code stores it in secret storage, not in chatLanguageModels.json
Spend limits in the docsSee spend management
keys / sk-agent-eval
capped
monthly cap$500 / $500
requests refused until the cap is raised
80% of the cap
email to the key owner
14 Apr
95% of the cap
webhook to #ai-spend
16 Apr
cap reached
key paused, other keys unaffected
17 Apr

The loop stopped at the ceiling, not at the invoice.

Questions

GitHub Copilot with Requesty, answered

Can I use my own API key with GitHub Copilot?

Yes. VS Code 1.122 and later support bring your own key through a Custom Endpoint provider. Add Requesty as the endpoint with your Requesty API key and the Messages API type, and Copilot Chat runs on the models you list.

Do I need a GitHub Copilot subscription to use BYOK?

Not for chat. Since VS Code 1.122, bring your own key works without a GitHub sign-in for chat, tools and MCP servers. Inline suggestions and next edit suggestions still need a GitHub sign-in.

Which models can I use in GitHub Copilot with Requesty?

Any model in the Requesty Model Library, including Claude, GPT, Gemini, DeepSeek and Kimi. Add each one to chatLanguageModels.json with its Requesty ID, for example anthropic/claude-sonnet-4-6 or openai/gpt-5.4, and the url https://router.requesty.ai.

How do I add more models to GitHub Copilot?

Run Chat: Manage Language Models from the Command Palette, select the gear next to the Requesty group and choose Add Model. The same menu has Update API Key for rotating your key.

Which API type should I choose for the Custom Endpoint?

Messages. It gives Anthropic models their native features, which the Chat Completions and Responses types do not carry. Set the url to https://router.requesty.ai without /v1, because VS Code appends /v1/messages itself.

How do I see GitHub Copilot token usage and cost?

Each Copilot Chat request is logged in the Requesty analytics dashboard with its model, tokens, cost and latency. Send a distinctive prompt once to confirm requests are arriving.

Why don't my Requesty models show up in the Copilot picker?

Run Chat: Manage Language Models from the Command Palette, re-add or re-select the Requesty Custom Endpoint, then reload the window. Check that the url is https://router.requesty.ai and the API key is valid.

Does BYOK cover commit messages and chat titles?

Those flows use a utility model. When VS Code shows the notification in the chat input, select Configure and set chat.utilityModel and chat.utilitySmallModel to one of your Requesty models.

Other agents

Guides for the other agents you run

The same key and the same dashboard work across every harness, so spend and logs stay in one place.

Claude CodeTwo environment variablesOpenCodeBuilt-in provider, run /connectHermes AgentNative plugin in the catalogPiOne extension, models load at startupKilo CodeBuilt-in provider in SettingsOpenClawBundled provider, one onboarding step
All coding agents on Requesty

Run GitHub Copilot on Requesty

600+ models, a spend cap on every key, and every session on the record.

Speak to founders