Integrations

AI providers

Connect your Claude, ChatGPT, Copilot, or Cursor subscription, or bring an API key from any provider in the catalog.

Velork runs Agent on the AI you connect. Sign in with a subscription you already pay for, or paste an API key. Credentials stay on this computer, in the system's secure storage.

Every account also gets Velork AI credits, so you can try Agent before you connect anything. See Velork AI credits.

Connect a subscription

These providers sign in with the account you already have. Connect them from the model picker in chat, during first-run setup, or by running the command from the Command Palette (Cmd + Shift + P).

In the model picker, each provider has a switch. Turn it on to connect, or off to disconnect.

Provider	Command	How it signs in
Claude Agent	Connect Claude Agent	Your Claude subscription, through the Claude Code login
OpenAI	Connect OpenAI (ChatGPT)	Your ChatGPT subscription in the browser, or an existing `codex login`
GitHub Copilot	Connect GitHub Copilot	GitHub device code
Cursor	Connect Cursor	Cursor User API key

Claude Agent

Run Connect Claude Agent
Sign in with your Claude account in the browser
Copy the code the page shows and paste it into Velork. Include the whole code, with the #.

To switch Claude accounts, run Connect Claude Agent again and choose Sign in again.

OpenAI (ChatGPT)

Run Connect OpenAI (ChatGPT)
Approve access in the browser. The window closes when it's done.

If you already ran codex login in a terminal, Velork picks that login up. OpenAI: Connection Status shows whether you're connected through the ChatGPT subscription or the CLI login.

GitHub Copilot

Run Connect GitHub Copilot
Click Copy Code & Open Browser
Paste the device code on GitHub and approve

Cursor

Create a User API key at cursor.com/dashboard/api
Run Connect Cursor and paste the key

Velork loads your Cursor models into the picker after you connect.

Add an API key

Open Settings → Models. You can add a key in two places.

Providers lists every provider in the Velork catalog. Search for one, click Connect, and fill in the credentials it asks for. Velork tests the credentials before it saves them, and shows an error instead of saving if they don't work.

Provider type	What Velork asks for
Most providers	An API key
AWS (Amazon Bedrock)	AWS region, access key ID, and secret access key. A Bedrock bearer token is optional.
Azure	Azure resource name and API key
Google Cloud (Vertex AI)	Google Cloud project, Vertex location, and service account JSON
Providers that need several values	One field for each environment variable the provider lists

Connected providers show Connected and a Manage button. Manage lets you reconnect with different credentials.

API keys is the collapsed Provider API keys section at the bottom of the page. It has quick fields for common providers:

Provider	What you need
OpenRouter	OpenRouter API key (`sk-or-...`)
GreenPT	GreenPT API key
OpenAI	OpenAI API key, for GPT models billed to your OpenAI account
Google AI	Google AI API key, for Gemini models
Custom (OpenAI-compatible)	API key and base URL of any OpenAI-compatible endpoint

OpenRouter and GreenPT are also in the model picker, and you can connect them with Connect OpenRouter and Connect GreenPT.

The full catalog is at the end of this page, with the environment variables each provider needs.

Choose models

Settings → Models → Your models lists the models from everything you've connected, grouped by provider. Search the list by model or provider name. Turn a model off to hide it from the chat picker. Click the refresh button to reload the list after you change a plan with your provider.

Pick the model for a chat from the model picker in the chat input.

Velork AI credits

Velork AI is a small set of models Velork hosts, so Agent works before you connect a provider. Requests on Velork AI use credits.

Plan	Credits
Hobby	2,000 once, when you start
Pro	A monthly allowance
Team	A larger monthly allowance, shared by the workspace

Current amounts are on pricing. Pro and Team also unlock larger Velork AI models.

When credits run out, chat shows Out of credits with a Connect AI button. Connect any provider above to keep going. Velork doesn't limit or charge for requests made with your own provider.

Disconnect

Disconnecting deletes the stored credentials. You can do it three ways:

Turn the provider's switch off in the model picker
Run the provider's disconnect command, for example Disconnect Claude Agent, Disconnect OpenAI, or Disconnect Cursor
For API keys, open Settings → Models and click the trash button next to the provider's model in Your models, or next to the key in Provider API keys

Troubleshooting

If chat fails with an auth error:

Run the provider's status command, such as GitHub Copilot: Connection Status
Connect again. A reconnect replaces the stored credentials.
Check that the key or subscription is still active with the provider

Usage from every provider is under Settings → Usage. See Plans and usage.

Provider catalog

114 providers supported in the Velork desktop app. Connect any of these with your own API keys or cloud credentials in Settings → Models → Providers.

Catalog updated 2026-07-31

Showing 114 of 114

302.AI

302ai

API key

Environment

302AI_API_KEY

SDK package

@ai-sdk/openai-compatible

Connect a subscription

Claude Agent

OpenAI (ChatGPT)

GitHub Copilot

Cursor

Add an API key

Choose models

Velork AI credits

Disconnect

Troubleshooting

Next steps

Agent overview

MCP servers

Plans and usage

Provider catalog

302.AI

Abacus

abliteration.ai

AIHubMix

Alibaba

Alibaba (China)

Alibaba Coding Plan

Alibaba Coding Plan (China)

Amazon Bedrock

Azure

Azure Cognitive Services

Bailing

Baseten

Berget.AI

Cerebras

Chutes

Clarifai

CloudFerro Sherlock

Cloudflare AI Gateway

Cloudflare Workers AI

Cohere

Cortecs

D.Run (China)

Databricks

Deep Infra

DeepSeek

DigitalOcean

DInference

evroc

FastRouter

Fireworks (Firepass)

Fireworks AI

Friendli

FrogBot

GitLab Duo

Groq

Helicone

HPC-AI

Hugging Face

iFlow

Inception

Inference

IO.NET

Jiekou.AI

Kilo Gateway

Kimi For Coding

KUAE Cloud Coding Plan

Llama

LLM Gateway

LMStudio

LucidQuery AI

Meganova

MiniMax (minimax.io)

MiniMax (minimaxi.com)

MiniMax Coding Plan (minimax.io)

MiniMax Coding Plan (minimaxi.com)

Mistral

Mixlayer

Moark

ModelScope

Moonshot AI

Moonshot AI (China)

Morph

NanoGPT

Nebius Token Factory