i2v.ai remote MCP server
Official metadata for the i2v.ai remote MCP server: discover models, quote inputs, generate media and track tasks from an MCP client. Streamable HTTP · OAuth mcp:read / mcp:write · USD ceiling on paid tasks.
Connect
https://i2v.ai/api/mcp
OAuth discovery: https://i2v.ai/.well-known/oauth-protected-resource/api/mcp
mcp:read— discover models, read parameters, quote inputs, check balance and usage, track your own tasks.mcp:write— create tasks and upload reference images. Paid tasks reserve and consume your USD balance.
Read is granted first; write needs an explicit second consent and can be revoked from the account settings.
Client guides for ChatGPT, Codex, [CC], Cursor, GitHub Copilot, VS Code and Gemini CLI live at https://i2v.ai/docs/mcp.
Server metadata
server.json is the source of truth for the Official MCP Registry entry io.github.i2v-ai/mcp. The registry.yml workflow validates the file and publishes it through GitHub OIDC.
./mcp-publisher validate server.json # after downloading the publisher binary
For AI agents
- Agent index: https://i2v.ai/llms.txt
- Full documentation: https://i2v.ai/llms-full.txt
- Skill discovery index: https://i2v.ai/.well-known/agent-skills/index.json
Related repositories
| Repository | Purpose |
|---|---|
| i2v-skills | Agent Skills and plugin manifests |
| i2v-cli | Command-line tools and local stdio MCP |
| i2v-sdk-js | TypeScript SDK |
| i2v-sdk-python | Python SDK |
| generate-action | GitHub Action |
| i2v-api-examples | Runnable examples |
| i2v-agent-tools | Ecosystem hub |
License
MIT. See LICENSE.