Inference API
Connect applications to ActGen using a dedicated API key. Published model files and models available for inference are listed separately.
1. Choose an available model
The model catalog lists models currently available to serve requests. An empty catalog or an unavailable response means inference is not ready. A key alone does not make a model available.
No model is available for inference right now. You can create an API key; check the catalog before adding funds for inference.
View model catalog2. Add prepaid funds
Open Billing to add credits and review your spending limit. Requests require available prepaid funds in the account that pays for your usage. Creating an API key is free.
Open Billing3. Create your API key
By signing in, you agree to each current ActGen agreement: Terms of Service (version 2026-09-20), Usage & Subscription Agreement (version 2026-09-20), Data Processing Addendum (version 2026-08-10), and Order Form (version 2026-09-20). ActGen records the accepted version and a tamper-evident hash of the exact agreement text applied to your action. How ActGen handles information is described in the Privacy Policy.
Request limits
Each user can send up to 30 requests per 60 seconds, with up to 2 requests running at once. Keys for the same user share these limits. Available funds and spending limits also apply.
Start with one request before increasing concurrency. A 429 response means a request limit was reached; a 402 response means funding is insufficient.
API addresses
- Base URL
- https://actgen.ai/api/v1
- Models
- GET /api/v1/models
- OpenRouter provider catalog
- GET /api/v1/models?format=openrouter
- Chat completions
- POST /api/v1/chat/completions
- Authentication header
- Authorization: Bearer YOUR_ACTGEN_API_KEY