GET /api/v1/inference/status/{request_id}

API version: v1.2.4

Check a MiniMax H3 Standard or Turbo generation request using the request_id returned by Submit Request. Text-to-video, image-to-video and reference-to-video use the same status endpoint. Other Model API generation requests also use this endpoint.

For model selection, see MiniMax H3, Standard text-to-video and Turbo text-to-video.

Endpoint

Plain text

GET /api/v1/inference/status/{request_id}

Include X-API-Key in the header and use the account that owns the request. Keep API keys on your application backend; do not put them in browser code or URL parameters.

Parameters

Parameter	Location	Type	Description
`request_id`	Path	string	Required. Preserve the ID returned by submission so you can resume checking the same request.
`logs`	Query	boolean	Optional, default `true`. Request available log entries; entries may still be absent. Use `false` for a smaller response.
`details`	Query	boolean	Optional, default `false`. Request additional task information when available. The basic workflow below does not require it.

The ID belongs in the path: /api/v1/inference/status/{request_id}. A query parameter does not replace it.

Response

Status available (200 OK)

An illustrative processing response:

JSON

{
  "status": "processing",
  "queue_position": 0,
  "request_id": "550e8400-e29b-41d4-a716-446655440000",
  "progress": 50,
  "logs": [
    {
      "message": "Task submitted successfully",
      "timestamp": "2026-10-21T10:30:00Z"
    }
  ]
}

Progress and logs are optional. A progress value is not a completion-time estimate.

Field	Type	Description
`request_id`	string	The submitted request's identifier.
`status`	string	`queued`, `processing`, `completed` or `failed` for a known request.
`queue_position`	integer	Queue indicator; `0` means no waiting position is reported, including terminal states.
`progress`	number	Optional progress indicator, when available. It can also appear in terminal responses.
`logs`	array	Optional entries with `message` and `timestamp`; do not require them to be present.
`error`	string	Optional error explanation. For terminal failure, check `status == "failed"`.
`result_url`	string	Optional completed-result preview URL. Use Get Generation Result for the media result.
`message`	string	Optional status explanation.
`details`	object	Optional additional information when requested and available.

Status updating (202 Accepted)

JSON

{
  "status": "processing",
  "queue_position": 0,
  "request_id": "550e8400-e29b-41d4-a716-446655440000",
  "message": "Task is being processed, status information is being updated"
}

No terminal outcome is available in this response. Preserve the same request_id and check again later.

What to do next

Status	Action
`queued`	Keep the request ID; the request is waiting.
`processing`	Keep the request ID; completion has not been confirmed.
`completed`	Fetch `/api/v1/inference/result/{request_id}`.
`failed`	Read the available `error`; stop waiting for this request. A new generation is a separate submission.
`not_found`	May accompany HTTP404. Check the ID and account; do not assume every404 body contains this field.

Examples

cURL

Set PHOSOR_API_KEY to your API key and REQUEST_ID to an existing request ID.

cURL

curl --fail-with-body --connect-timeout 10 --max-time 40 \
  "https://phosor.ai/api/v1/inference/status/${REQUEST_ID}?logs=false" \
  -H "X-API-Key: ${PHOSOR_API_KEY}"

Python: one status check

This function checks an existing request once. The timeout bounds each connection/read wait; it is not a generation deadline. Call it from your backend when you need an update.

Python

import requests


def check_status(api_key, request_id):
    response = requests.get(
        f"https://phosor.ai/api/v1/inference/status/{request_id}",
        headers={"X-API-Key": api_key},
        params={"logs": False},
        timeout=(10, 40),
    )
    response.raise_for_status()
    return response.json()  # Both 200 and transitional 202 are handled.

JavaScript: one status check

Use this on your backend. HTTP failures throw; they are not converted into a generation failure.

JavaScript

async function checkStatus(apiKey, requestId) {
  const response = await fetch(
    `https://phosor.ai/api/v1/inference/status/${encodeURIComponent(requestId)}?logs=false`,
    {
      headers: { 'X-API-Key': apiKey },
      signal: AbortSignal.timeout(40000),
    },
  );
  if (!response.ok) throw new Error(`Status lookup HTTP ${response.status}`);
  return response.json();
}

HTTP errors and recovery

HTTP response	Meaning and action
`401`	Authentication is required or invalid. Check the API key.
`403`	Access is forbidden. Use the account that owns this request.
`404`	The request cannot be found. Check the ID and account; it may no longer be available.
`502`	Status lookup failed. This does not establish that generation failed. Retry the lookup later using the same ID.
Network timeout	The lookup outcome is unknown. Retain the ID and resume checking it; do not automatically submit another generation.

Error bodies may differ; inspect the HTTP code before reading status fields. For example, an ownership check can return {"error":"Not found"} without status or request_id. This reference does not promise a fixed task-retention period.

Quick Start — select H3 Standard or Turbo and submit a request.
Submit Request — model-specific generation parameters.
Get Generation Result — retrieve and download completed media.
Video Limits — current input and output requirements.

GET /api/v1/inference/status/{request_id}

Endpoint

Parameters

Response

Status available (200 OK)

Status updating (202 Accepted)

What to do next

Examples

cURL

Python: one status check

JavaScript: one status check

HTTP errors and recovery

Related resources