API reference

Moton API

An OpenAI-compatible API for uncensored inference and dedicated GPU compute. Point any OpenAI client at Moton and start generating.

base https://console.moton.io/apiv1 · OpenAI-compatible

Quickstart

Moton is an OpenAI-compatible inference API. If you can call the OpenAI API, you can call Moton — most clients only need a change to the base URL and the API key.

1. Create an API key

In the dashboard, open API keys and click Create key. You will be shown the full key once. Store it — only a mt_ prefix is visible afterwards.

2. Make your first request

Every request sends your key in the Authorization header. The endpoint is /v1/chat/completions.

bash

curl https://console.moton.io/api/v1/chat/completions \
  -H "Authorization: Bearer mt_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3.8-flash-next-uncensored",
    "messages": [
      { "role": "user", "content": "Why is the sky blue?" }
    ]
  }'

The response is a standard chat completion:

json

{
  "id": "chatcmpl-9f2a1c",
  "object": "chat.completion",
  "model": "qwen3.8-flash-next-uncensored",
  "choices": [
    {
      "index": 0,
      "message": { "role": "assistant", "content": "It's Rayleigh scattering…" },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 48,
    "total_tokens": 60,
    "cost": 0.0004
  }
}

3. Try the models list

GET /v1/models returns every model on the platform. It does not require authentication.

Live example

Click “Run it” to call GET /api/v1/models against this instance.

Authentication

Authenticate by sending your API key in the Authorization header using the Bearer scheme.

bash

curl https://console.moton.io/api/v1/chat/completions \
  -H "Authorization: Bearer mt_..." \
  -d '{ "model": "qwen3.8-flash-next-uncensored", "messages": [] }'

Keys & permissions

Property	Value
Format	`mt_` prefix, 32 random characters
Visibility	The full secret is shown once at creation. Only the prefix is stored for display.
Scope	Per-account. Each key is billed to the account balance and honors its optional spend cap.
Rotation	Create a new key, update your clients, then revoke the old one.

Errors

If the key is missing or invalid, requests return 401:

json

{
  "error": {
    "message": "Invalid API key. Create one in the dashboard.",
    "type": "auth_error"
  }
}

Models

Moton hosts uncensored open-weights models. You select one by its id in the request body. Use GET /v1/models to list what is currently available.

Model ID	Organization	Context	Capabilities
`qwen3.8-flash-next-uncensored`	Moton	1M	tools, reasoning, json, vision
`qwen3.8-27b-uncensored`	Moton	256K	tools, reasoning, json, vision
`gpt-oss-20b`	OpenAI	128K	tools, reasoning, json
`qwen3.6-35b-a3b`	Alibaba	256K	tools, reasoning, json

Pricing is per-token and varies by model. Exact rates are shown live in the dashboard and on the pricing page.

Chat Completions

POST`/v1/chat/completions`

The core endpoint. It accepts a chat message history and returns either a single completion or a server-sent event stream.

Request body

Field	Type	Required	Description
`model`	string	required	A catalogue `id` (see `GET /v1/models`) or the name of one of your deployments.
`messages`	array	required	Chat history. Each item has a `role` of `system`, `user`, or `assistant` and a `content` string.
`stream`	boolean	optional	If `true`, stream the response as SSE chunks.
`temperature`	number	optional	Sampling temperature, 0.0 (deterministic) to 1.0 (creative).
`max_tokens`	integer	optional	Cap on generated tokens for this completion.

Non-streaming response

json

{
  "id": "chatcmpl-9f2a1c",
  "object": "chat.completion",
  "created": 1753250000,
  "model": "qwen3.8-flash-next-uncensored",
  "choices": [
    {
      "index": 0,
      "message": { "role": "assistant", "content": "It's Rayleigh scattering…" },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 48,
    "total_tokens": 60,
    "cost": 0.0004
  }
}

Streaming (SSE)

Set "stream": true. The response is a sequence of data: events ending with a terminal [DONE]. The final chunk carries token usage and cost.

text

data: {"id":"chatcmpl-9f2a1c","object":"chat.completion.chunk","model":"qwen3.8-flash-next-uncensored","choices":[{"index":0,"delta":{"content":"It"},"finish_reason":null}]}

data: {"id":"chatcmpl-9f2a1c","object":"chat.completion.chunk","model":"qwen3.8-flash-next-uncensored","choices":[{"index":0,"delta":{"content":"'s"},"finish_reason":null}]}

data: {"id":"chatcmpl-9f2a1c","object":"chat.completion.chunk","model":"qwen3.8-flash-next-uncensored","choices":[{"index":0,"delta":{},"finish_reason":"stop"}],"usage":{"prompt_tokens":12,"completion_tokens":48,"total_tokens":60,"cost":0.0004}}

data: [DONE]

Errors

Status	Meaning
`400`	Bad request — missing `model`, empty or missing `messages`, or a prompt past the model's context window.
`401`	Missing or invalid API key.
`402`	`insufficient_credits` — balance too low for serverless inference, or the key's spend cap has been reached.
`403`	`not_in_plan` — the model isn't on your subscription and you have no balance to fall back to per-token.
`404`	Unknown model `id` or deployment name.
`503`	`model_not_online` — the model is in the catalogue but no shared capacity is serving it yet.

Models List

GET`/v1/models`

Returns the list of models currently served. No authentication required.

json

{
  "object": "list",
  "data": [
    { "id": "qwen3.8-flash-next-uncensored", "object": "model", "created": 1755820800, "owned_by": "Moton" },
    { "id": "qwen3.8-27b-uncensored", "object": "model", "created": 1755820800, "owned_by": "Moton" },
    { "id": "gpt-oss-20b", "object": "model", "created": 1755820800, "owned_by": "OpenAI" },
    { "id": "qwen3.6-35b-a3b", "object": "model", "created": 1755820800, "owned_by": "Alibaba" }
  ]
}

Errors

Errors use the OpenAI error envelope. The type field is machine-readable; the message field is human-readable.

json

{
  "error": {
    "message": "Unknown model 'gpt-4'.",
    "type": "invalid_request_error"
  }
}

Type	Status	When it happens
`invalid_request_error`	`400`	Malformed body, empty messages, or unknown model id (404).
`auth_error`	`401`	Missing, expired, or invalid API key.
`billing_error`	`402`	Insufficient balance for serverless inference, or the key's spend cap reached.
`server_error`	`5xx`	Upstream model server error, or no shared capacity online (503).

Billing

Billing is usage-based. Inference is billed per token and privacy servers are billed per hour of runtime. Your account holds a balance, and every request accrues cost against it.

Where cost appears

Every completion response includes a usage object with a cost field (in USD). Streaming requests report total usage in the final chunk.

json

"usage": {
  "prompt_tokens": 12,
  "completion_tokens": 48,
  "total_tokens": 60,
  "cost": 0.0004
}

Balance & caps

Item	Behavior
Balance	Top up from the `Billing` page. Requests are billed against it in real time.
Spend cap	Optional per-key limit. When reached, the key stops accepting requests with `402`.
Invoices	Itemized monthly invoices are available on the `Billing` page.

New accounts receive $1.00 in free credit with no card required — enough to test the API before committing.

Changelog

Notes on recent changes to the API and platform.

Date	Change
2026-08	Initial public API. OpenAI-compatible chat completions + model list.
2026-08	Per-key spend caps and live cost meter in the playground.

Moton API

Quickstart

1. Create an API key

2. Make your first request

3. Try the models list

Authentication

Keys & permissions

Errors

Models

Chat Completions

POST/v1/chat/completions

Request body

Non-streaming response

Streaming (SSE)

Errors

Models List

GET/v1/models

Errors

Billing

Where cost appears

Balance & caps

Changelog

POST`/v1/chat/completions`

GET`/v1/models`