API reference

Public API reference.

Use this reference to choose the right endpoint and avoid assuming resource APIs that are not available in the current public API.

Start quickstart View models

Request shape

Every endpoint starts from the same base URL and bearer key.

Pick a model id from Models and confirm prices in Pricing before sustained traffic.

HTTP shape

POST https://api.1api.club/v1/chat/completions
Authorization: Bearer $API_KEY
Content-Type: application/json

Reference scope

This page describes public API behavior, endpoint availability, and client-visible response shapes.

Base URL: https://api.1api.club/v1
Authentication: Authorization: Bearer $API_KEY
Response format: OpenAI-compatible JSON or SSE where the endpoint supports streaming.
Request id: Responses include SDK-readable request id headers for support and debugging.

Compatibility baseline

SDK types can be wider than the public API.

Public support is split between ready, limited, and unavailable workloads. Cross-request resource object lifecycle is excluded unless this page says otherwise.

Baseline

Endpoint visibility follows the current reference target, but every limited endpoint still depends on model capability and account availability.

Supported endpoints

Use the endpoint that matches the workload.

These endpoints are the public OpenAI-compatible surface to check first. Limited rows are real surface area, but not a blanket promise for every model.

Ready

Ready when the selected public model is available for the workload.

Limited

Available only when the selected model and account support that workload.

Not available

Not part of the current public API. Use documented errors as the contract.

Method: GET
Path: /v1/models
Workload: Models
Status: Ready
Notes: List public model ids available to clients.

Method: GET
Path: /v1/models/{model}
Workload: Models
Status: Ready
Notes: Retrieve one public model record with client-visible availability details.

Method: POST
Path: /v1/chat/completions
Workload: Text
Status: Ready
Notes: Chat Completions JSON and streaming baseline for common clients.

Method: POST
Path: /v1/completions
Workload: Text
Status: Limited
Notes: Legacy Completions, including n, seed, and non-streaming best_of where the selected model supports them.

Method: POST
Path: /v1/responses
Workload: Text
Status: Ready
Notes: Responses JSON and streaming baseline with supported continuation behavior.

Method: POST
Path: /v1/responses/compact
Workload: Text
Status: Limited
Notes: Compact Responses request path when compact handling is configured for the model.

Method: GET
Path: /v1/responses
Workload: Text
Status: Limited
Notes: Responses WebSocket runtime only for WebSocket upgrade requests; normal HTTP GET is not a list endpoint.

Method: GET
Path: /v1/realtime
Workload: Realtime
Status: Limited
Notes: Realtime WebSocket relay when the selected realtime model and route are configured.

Method: POST
Path: /v1/responses/input_tokens
Workload: Text
Status: Limited
Notes: Input token count helper when the selected model and route expose this endpoint.

Method: POST
Path: /v1/embeddings
Workload: Embeddings
Status: Limited
Notes: Embeddings baseline when a public model is configured for embeddings.

Method: POST
Path: /v1/moderations
Workload: Safety
Status: Limited
Notes: Moderation baseline with no billable usage charge when a moderation model is configured.

Method: POST
Path: /v1/images/generations
Workload: Images
Status: Ready
Notes: Image generation workload, including streaming where the selected model supports it.

Method: POST
Path: /v1/images/edits
Workload: Images
Status: Ready
Notes: Image edit workload with multipart intake and streaming where supported.

Method: POST
Path: /v1/videos
Workload: Videos
Status: Limited
Notes: Create a video generation task when a video-capable public model is configured.

Method: GET
Path: /v1/videos/{video_id}
Workload: Videos
Status: Limited
Notes: Retrieve a video task by the returned video id from create or remix.

Method: GET
Path: /v1/videos/{video_id}/content
Workload: Videos
Status: Limited
Notes: Download generated video content or a supported variant after the video task is ready.

Method: POST
Path: /v1/videos/{video_id}/remix
Workload: Videos
Status: Limited
Notes: Create a remix task from a returned video id from create or retrieve.

Method: POST
Path: /v1/audio/transcriptions
Workload: Audio
Status: Limited
Notes: Non-streaming transcription baseline when audio REST is enabled for the selected model.

Method: POST
Path: /v1/audio/translations
Workload: Audio
Status: Limited
Notes: Non-streaming translation baseline when audio REST is enabled for the selected model.

Method: POST
Path: /v1/audio/speech
Workload: Audio
Status: Limited
Notes: Speech binary passthrough when audio REST is enabled for the selected model.

Supported public OpenAI-compatible endpoints
Method	Path	Workload	Status	Notes
GET	/v1/models	Models	Ready	List public model ids available to clients.
GET	/v1/models/{model}	Models	Ready	Retrieve one public model record with client-visible availability details.
POST	/v1/chat/completions	Text	Ready	Chat Completions JSON and streaming baseline for common clients.
POST	/v1/completions	Text	Limited	Legacy Completions, including n, seed, and non-streaming best_of where the selected model supports them.
POST	/v1/responses	Text	Ready	Responses JSON and streaming baseline with supported continuation behavior.
POST	/v1/responses/compact	Text	Limited	Compact Responses request path when compact handling is configured for the model.
GET	/v1/responses	Text	Limited	Responses WebSocket runtime only for WebSocket upgrade requests; normal HTTP GET is not a list endpoint.
GET	/v1/realtime	Realtime	Limited	Realtime WebSocket relay when the selected realtime model and route are configured.
POST	/v1/responses/input_tokens	Text	Limited	Input token count helper when the selected model and route expose this endpoint.
POST	/v1/embeddings	Embeddings	Limited	Embeddings baseline when a public model is configured for embeddings.
POST	/v1/moderations	Safety	Limited	Moderation baseline with no billable usage charge when a moderation model is configured.
POST	/v1/images/generations	Images	Ready	Image generation workload, including streaming where the selected model supports it.
POST	/v1/images/edits	Images	Ready	Image edit workload with multipart intake and streaming where supported.
POST	/v1/videos	Videos	Limited	Create a video generation task when a video-capable public model is configured.
GET	/v1/videos/{video_id}	Videos	Limited	Retrieve a video task by the returned video id from create or remix.
GET	/v1/videos/{video_id}/content	Videos	Limited	Download generated video content or a supported variant after the video task is ready.
POST	/v1/videos/{video_id}/remix	Videos	Limited	Create a remix task from a returned video id from create or retrieve.
POST	/v1/audio/transcriptions	Audio	Limited	Non-streaming transcription baseline when audio REST is enabled for the selected model.
POST	/v1/audio/translations	Audio	Limited	Non-streaming translation baseline when audio REST is enabled for the selected model.
POST	/v1/audio/speech	Audio	Limited	Speech binary passthrough when audio REST is enabled for the selected model.

Compatibility helpers

Some compatibility behavior is intentionally bounded.

These helpers exist for client compatibility. They are narrower than the full resource APIs exposed by general SDK types.

Responses object reads: Completed response retrieve and completed-stream replay are bounded compatibility helpers, not a full cross-request object lifecycle promise.
Video task ids: Video retrieve, content download, and remix use returned video ids to continue task operations.
Endpoint status: Limited endpoints require a matching public model and account availability.

Not available

Do not build on unavailable resource APIs.

These surfaces are not promised by the current public API. Treat shared SDK type availability as client library surface, not product support.

Responses lifecycle: Background mode, full object lifecycle, cancel semantics beyond the bounded local helper, and list APIs.
Resource storage: Files, uploads, vector stores, and batches.
DALL-E legacy: DALL-E-only image variations, POST /v1/images/variations, and legacy DALL-E model surfaces.
Videos resource management: Video list, delete, characters, edits, extensions, and legacy video generation routes are not part of the current public API.
Managed resources: Fine-tuning, evals, assistants, threads, containers, conversations, and beta resources.
Admin APIs: OpenAI Admin organization, project, usage, and audit log APIs.
Stored Chat: Stored Chat objects and cross-request Chat message retrieval.

Safety

Errors and model data stay client-safe.

The reference intentionally describes only behavior that clients can rely on.

Unavailable resources: Unavailable resources return documented errors.
Client-safe errors: Internal failure details and backend-only identifiers are not exposed to clients.
Model data: Public model responses only include client-visible model details.

Turn endpoint choice into an operational check.

For production use, combine this reference with model availability, public pricing, usage review, and trust checks.

Model ids

Public model catalog.

Spend

Pricing, usage, and credits.

Health

Public trust context and support details.

Pages

Continue through the docs.

Use the overview for integration context or quickstart for a first request.

OverviewConnection facts, public scope, and the first verification path.Open QuickstartCreate a key, set the base URL, and send the first request.Open API referenceSupported endpoints, request shape, and unavailable resources.Current