# MyInferenceAPI

> Fast, OpenAI-compatible LLM inference

Base URL: https://api.myinferenceapi.com

Authentication: `Authorization: Bearer <your_api_key>`

# MyInferenceAPI

OpenAI-compatible LLM inference for agentic, tool-calling, and coding workloads.

## Endpoints
- POST /v1/chat/completions — OpenAI-compatible chat completions. Supports streaming, JSON mode, structured outputs, and tool calling.
- GET /v1/models — List available models.

## Models
- Qwen/Qwen3-Coder-30B-A3B-Instruct — 30B mixture-of-experts coder model, 64K context. Tool calling + structured outputs.

## Authentication
Bearer token in the Authorization header.


## Agent Onboarding

To use this or any other MyAPIHQ service as an autonomous agent, follow these steps:

1. **Register an Agent Account**:
   Create an agent account. Returns `account_id`, `pin`, and a `token` (JWT) in one call — no email or human required.
   ```bash
   curl -X POST https://api.myapihq.com/hq/account/agent/create \
     -H "Content-Type: application/json" \
     -d '{"label": "my-agent"}'
   ```

2. **Create a Permanent API Key**:
   Use the `token` from step 1 to create a workspace-scoped API key (`hq_live_...`).
   ```bash
   curl -X POST https://api.myapihq.com/hq/account/create/key \
     -H "Authorization: Bearer <token_from_step_1>" \
     -H "Content-Type: application/json" \
     -d '{"name": "my-agent-key"}'
   ```

3. **Authenticate**:
   All subsequent requests must include: `Authorization: Bearer <your_api_key>`.
   The JWT token from step 1 is valid for account and billing endpoints only.
   All service calls (orgs, funnels, email, domains) require the API key from step 2.

## Endpoints

### POST /v1/chat/completions
**Create a chat completion**


**Request Body (JSON):**
```json
{
  "model": string - Model id, e.g. Qwen/Qwen3-Coder-30B-A3B-Instruct. (Required),
  "messages": Array<object> (Required),
  "max_tokens": integer - Maximum tokens to generate.,
  "temperature": number,
  "top_p": number,
  "stream": boolean - Stream the response as server-sent events.,
  "tools": Array<object>,
  "response_format": object - Set to {"type":"json_object"} for JSON mode.
}
```

**Response (JSON):**
```json
{
  "id": string,
  "object": string,
  "created": integer,
  "model": string,
  "choices": Array<object>,
  "usage": object - Prompt, completion, and total token counts.
}
```


### GET /v1/models
**List available models**


**Response (JSON):**
```json
{
  "object": string,
  "data": Array<{
    "id": string,
    "object": string,
    "created": integer,
    "owned_by": string
  }>
}
```


## Pricing

Pay per use. No subscription.

---
## Ecosystem Integration
This service is part of the MyAPIHQ ecosystem. You authenticate using an API Key generated from your MyAPIHQ account.
To view the full list of available services (like domain registration, funnel building, transactional email, object storage, etc.), fetch [https://myapihq.com/llms.txt](https://myapihq.com/llms.txt).