❱ API Reference

Base URL:

https://inference-api-611798501438.us-central1.run.app

The Majestix AI Inference Hub API is a multi-model AI gateway that provides unified access to models from OpenAI, Anthropic, Google Vertex AI, and OpenRouter through a single credits-based billing system.

---

Authentication

All endpoints require authentication via one of two methods:

Method	Header(s)	Use Case
API Key	X-Api-Key: inf_...	VSCode extension, CLI tools, programmatic access
Firebase Auth + App Check	Authorization: Bearer <id_token> and X-Firebase-AppCheck: <token>	Browser-based web applications

API keys are issued through the /api-keys endpoint and use the inf_ prefix. Firebase authentication is handled through the web SDK and is intended exclusively for browser clients.

---

Streaming Protocol (SSE)

Streaming endpoints (/chat, /code) return Server-Sent Events (SSE) with the following event types:

Event Type	Description
chunk	Incremental text content: {"type": "chunk", "content": "..."}
tool_use	Tool call from the model: {"type": "tool_use", "id": "...", "name": "...", "input": {...}}
done	Stream complete: {"type": "done", "model": "...", "session_id": "...", "credits_used": 42}
error	Error during generation: {"type": "error", "message": "..."}

Each SSE message is formatted as:

data: {"type": "chunk", "content": "Hello"}\n\n

---

Endpoints

Method	Path	Description
POST	/chat	Chat completions (streaming and non-streaming)
POST	/code	Agentic tool-use endpoint for coding assistants
GET	/models	List available models and capabilities
GET	/session/{id}/history	Retrieve conversation history
DELETE	/session/{id}	Delete a session
GET	/usage/me	Credit usage and plan information
POST	/api-keys	Generate a new API key
GET	/api-keys	List existing API keys
DELETE	/api-keys/{id}	Revoke an API key
POST	/billing/checkout	Create Stripe subscription checkout
POST	/billing/topup	One-time credit top-up purchase
POST	/billing/portal	Open Stripe customer portal
POST	/billing/sync-balance	Force credit balance cache refresh
GET	/admin/usage	Platform-wide analytics (admin only)
GET	/admin/model-economics	Per-model cost and revenue breakdown (admin only)

---

Credits

All API usage is billed in credits. 1 credit equals $0.001 USD. A 1.4x margin is applied to raw provider costs. Credits are reserved at the worst-case estimate before generation begins and reconciled to the actual amount when the response completes.

Plan	Monthly Price	Credits Included
Free	$0	500
Guru	$10/mo	10,000
Pro	$50/mo	55,000

Additional credits can be purchased as one-time top-ups through the /billing/topup endpoint.

---

Error Responses

All endpoints return errors in a consistent JSON format:

{
  "detail": "Insufficient credits"
}

Status Code	Meaning
400	Bad request (invalid parameters)
401	Authentication failed or missing
403	Forbidden (insufficient permissions or credits)
404	Resource not found
429	Rate limit exceeded
500	Internal server error