### GET /replays

Source: https://aigateway.sh/openapi.json

API endpoint for GET /replays

```markdown
### Responses

#### 200 - Replay list

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/replays"
```

```

--------------------------------

### GET /files

Source: https://aigateway.sh/openapi.json

API endpoint for GET /files

```markdown
### Parameters

- **purpose** (string, query, optional)

### Responses

#### 200 - File list

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/files?purpose=string"
```

```

--------------------------------

### GET /models

Source: https://aigateway.sh/openapi.json

API endpoint for GET /models

```markdown
### Parameters

- **modality** (string (text|image|audio-stt|audio-tts|video|embedding|rerank|moderation), query, optional)
- **provider** (string, query, optional) (example: "anthropic")

### Responses

#### 200 - OpenAI-shaped model list with pricing + capabilities

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/models?modality=text&provider=anthropic"
```

```

--------------------------------

### GET /files/jobs/{jobId}/{filename}/signed

Source: https://aigateway.sh/openapi.json

Returns a publicly fetchable URL on media.aigateway.sh, valid until expires_at. Max expiry 7 days.

```markdown
### Parameters

- **jobId** (string, path, required)
- **filename** (string, path, required)
- **expires_in** (integer, query, optional)

### Responses

#### 200 - Signed URL

- **url** (string)
- **expires_at** (integer)

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/files/jobs/{jobId}/{filename}/signed?expires_in=3600"
```

```

--------------------------------

### GET /balance

Source: https://aigateway.sh/openapi.json

API endpoint for GET /balance

```markdown
### Responses

#### 200 - Balance

- **cents** (integer)
- **usd** (number)

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/balance"
```

```

--------------------------------

### GET /replays/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /replays/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Replay record

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/replays/{id}"
```

```

--------------------------------

### GET /files/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /files/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - File metadata

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/files/{id}"
```

```

--------------------------------

### GET /realtime

Source: https://aigateway.sh/openapi.json

WebSocket endpoint (connect with wss://). URL: wss://api.aigateway.sh/v1/realtime?model=deepgram/nova-3&encoding=linear16&sample_rate=16000&interim_results=true. Auth via ?api_key= (browsers) or Authorization: Bearer (servers). Stream raw audio frames; receive {type:"Results",...} messages then a final {type:"Metadata"}. End with {"type":"CloseStream"}. Billed per audio-minute. Models: deepgram/nova-3, deepgram/flux.

```markdown
### Responses

#### 101 - Switching Protocols — WebSocket established

Switching Protocols — WebSocket established

#### 426 - Upgrade Required — connect with wss://

Upgrade Required — connect with wss://

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/realtime"
```

```

--------------------------------

### GET /batches

Source: https://aigateway.sh/openapi.json

API endpoint for GET /batches

```markdown
### Parameters

- **limit** (integer, query, optional)

### Responses

#### 200 - Batch list

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/batches?limit=20"
```

```

--------------------------------

### POST /audio/transcriptions

Source: https://aigateway.sh/openapi.json

Synchronous by default. Set async:true (then poll GET /v1/jobs/{id}) or pass a webhook_url to run as a batch job — best for long recordings (Deepgram Nova 3 / Flux). For live transcription, open the WebSocket at /v1/realtime.

```markdown
### Request Body

**Content-Type:** multipart/form-data

- **model** (string) (required) (example: "deepgram/nova-3")
- **file** (string (binary)) (required)
- **language** (string)
- **async** (boolean): Run as an async job; returns 202 with a job id.
- **webhook_url** (string (uri)): POST the signed result here when ready (implies async).

**Content-Type:** application/json

- **model** (string) (required) (example: "deepgram/nova-3")
- **audio_url** (string (uri)) (required)
- **language** (string)
- **async** (boolean): Run as an async job; returns 202 with a job id.
- **webhook_url** (string (uri)): POST the signed result here when ready (implies async).

### Responses

#### 200 - Transcription result (sync)

Empty response body

#### 202 - Accepted — async job created; poll GET /v1/jobs/{id} or await the webhook

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/audio/transcriptions" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepgram/nova-3",
  "file": "string",
  "language": "string",
  "async": "true",
  "webhook_url": "string"
}'
```

```

--------------------------------

### GET /usage/by-tag

Source: https://aigateway.sh/openapi.json

API endpoint for GET /usage/by-tag

```markdown
### Parameters

- **month** (string, query, optional) (example: "2026-04")

### Responses

#### 200 - Aggregated usage per tag

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/usage/by-tag?month=2026-04"
```

```

--------------------------------

### GET /files/{id}/content

Source: https://aigateway.sh/openapi.json

API endpoint for GET /files/{id}/content

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Raw file bytes

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/files/{id}/content"
```

```

--------------------------------

### GET /usage/by-sub-account

Source: https://aigateway.sh/openapi.json

API endpoint for GET /usage/by-sub-account

```markdown
### Parameters

- **month** (string, query, optional) (example: "2026-04")

### Responses

#### 200 - Aggregated usage per customer

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/usage/by-sub-account?month=2026-04"
```

```

--------------------------------

### POST /videos/generations

Source: https://aigateway.sh/openapi.json

Returns 202 with a job id. Poll GET /jobs/{id} or pass webhook_url.

```markdown
### Request Body

**Content-Type:** application/json

- **prompt** (string) (required)
- **model** (string)
- **duration** (number)
- **aspect_ratio** (string)
- **resolution** (string)
- **image_url** (string (uri))
- **webhook_url** (string (uri))

### Responses

#### 202 - Job accepted

**JobAccepted**
- **id** (string) (required) (example: "job_abc123")
- **status** (string (queued)) (required) ("queued")
- **object** (string (job)) (required) ("job")

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/videos/generations" \
  -H "Content-Type: application/json" \
  -d '{
  "prompt": "string",
  "model": "string",
  "duration": "0",
  "aspect_ratio": "string",
  "resolution": "string",
  "image_url": "string",
  "webhook_url": "string"
}'
```

```

--------------------------------

### GET /models/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /models/{id}

```markdown
### Parameters

- **id** (string, path, required) (example: "anthropic/claude-opus-4.7")

### Responses

#### 200 - Single-model record

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/models/{id}"
```

```

--------------------------------

### GET /evals

Source: https://aigateway.sh/openapi.json

API endpoint for GET /evals

```markdown
### Responses

#### 200 - Eval list

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/evals"
```

```

--------------------------------

### GET /sub-accounts

Source: https://aigateway.sh/openapi.json

API endpoint for GET /sub-accounts

```markdown
### Responses

#### 200 - Sub-account list

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/sub-accounts"
```

```

--------------------------------

### GET /jobs/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /jobs/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Job status + result

**Job**
- **id** (string) (required)
- **status** (string (queued|processing|completed|failed)) (required) ("queued"|"processing"|"completed"|"failed")
- **modality** (string)
- **model** (string)
- **created_at** (integer)
- **updated_at** (integer)
- **error** (object)
  - **message** (string)
  - **code** (string)
- **result_file_id** (string)
- **result_url** (string (uri))
- **webhook_url** (string (uri))
- **webhook_delivered** (boolean)
- **attempts** (integer)

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/jobs/{id}"
```

```

--------------------------------

### GET /health/providers

Source: https://aigateway.sh/openapi.json

API endpoint for GET /health/providers

```markdown
### Responses

#### 200 - p50, p95, error rate per upstream

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/health/providers"
```

```

--------------------------------

### GET /batches/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /batches/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Batch record

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/batches/{id}"
```

```

--------------------------------

### GET /webhook-secret

Source: https://aigateway.sh/openapi.json

API endpoint for GET /webhook-secret

```markdown
### Responses

#### 200 - Signing secret

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/webhook-secret"
```

```

--------------------------------

### GET /sub-accounts/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /sub-accounts/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Sub-account record

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/sub-accounts/{id}"
```

```

--------------------------------

### GET /evals/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for GET /evals/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Eval result + winner_model

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/evals/{id}"
```

```

--------------------------------

### GET /sub-accounts/{id}/usage

Source: https://aigateway.sh/openapi.json

API endpoint for GET /sub-accounts/{id}/usage

```markdown
### Parameters

- **id** (string, path, required)
- **month** (string, query, optional) (example: "2026-04")

### Responses

#### 200 - Usage report

Empty response body

### Example Usage

```bash
curl -X GET "https://api.aigateway.sh/v1/sub-accounts/{id}/usage?month=2026-04"
```

```

--------------------------------

### POST /chat/completions

Source: https://aigateway.sh/openapi.json

Drop-in replacement for OpenAI's /v1/chat/completions. Supports streaming via SSE, tool calling, vision, JSON mode, and reasoning models with normalized `reasoning_content` deltas. Set `model:"auto"` (or `auto/text`, or omit it) to use the Auto Router; bias the pick with the `x-routing` header and cap cost with `baseline_model`. Routed responses return `X-Routing-*` and `X-Auto-*` headers showing the selected model and dollars saved.

```markdown
### Request Body

**Content-Type:** application/json

- **model** (string) (required): Provider-qualified slug, or "auto" / "auto/<modality>" to let the Auto Router pick the cheapest model in a curated pool that clears the quality floor (never above your baseline). Featured: anthropic/claude-opus-4.8, minimax/m3, google/gemini-3.1-pro, anthropic/claude-sonnet-4.6, openai/gpt-5.5, google/gemini-2.5-pro, moonshot/kimi-k2.6, xai/grok-4.20-multi-agent-0309, openai/gpt-5.4-mini, alibaba/qwen3.5-397b-a17b. Fetch full list from /models. Catalog: 1000+. (example: "anthropic/claude-opus-4.8")
- **baseline_model** (string): Auto Router only: the model you'd otherwise call. The router routes DOWN from it, so it doubles as a hard cost ceiling — you never pay more than calling it directly. Defaults to the premium model for the modality. Used only when model is "auto" / "auto/<modality>" or omitted. (example: "anthropic/claude-opus-4.8")
- **messages** (array (object)) (required)
  Array items:
    - **role** (string (system|user|assistant|tool)) (required) ("system"|"user"|"assistant"|"tool")
    - **content** (string) (required)
- **stream** (boolean)
- **temperature** (number)
- **max_tokens** (integer)
- **tools** (array)
- **tool_choice** (string)
- **response_format** (object)
- **webhook_url** (string (uri)): If set, delivery flips to async-job mode (202 + signed callback)

### Responses

#### 200 - Chat completion (SSE if stream=true)

Empty response body

#### 400 - response

- **error** (object) (required)
  - **message** (string) (required)
  - **type** (string (invalid_request_error|authentication_error|budget_exceeded|model_not_found|rate_limit_error|provider_error|timeout_error)) (required) ("invalid_request_error"|"authentication_error"|"budget_exceeded"|"model_not_found"|"rate_limit_error"|"provider_error"|"timeout_error")
  - **code** (integer)

#### 401 - response

- **error** (object) (required)
  - **message** (string) (required)
  - **type** (string (invalid_request_error|authentication_error|budget_exceeded|model_not_found|rate_limit_error|provider_error|timeout_error)) (required) ("invalid_request_error"|"authentication_error"|"budget_exceeded"|"model_not_found"|"rate_limit_error"|"provider_error"|"timeout_error")
  - **code** (integer)

#### 402 - response

- **error** (object) (required)
  - **message** (string) (required)
  - **type** (string (invalid_request_error|authentication_error|budget_exceeded|model_not_found|rate_limit_error|provider_error|timeout_error)) (required) ("invalid_request_error"|"authentication_error"|"budget_exceeded"|"model_not_found"|"rate_limit_error"|"provider_error"|"timeout_error")
  - **code** (integer)

#### 404 - response

- **error** (object) (required)
  - **message** (string) (required)
  - **type** (string (invalid_request_error|authentication_error|budget_exceeded|model_not_found|rate_limit_error|provider_error|timeout_error)) (required) ("invalid_request_error"|"authentication_error"|"budget_exceeded"|"model_not_found"|"rate_limit_error"|"provider_error"|"timeout_error")
  - **code** (integer)

#### 429 - response

- **error** (object) (required)
  - **message** (string) (required)
  - **type** (string (invalid_request_error|authentication_error|budget_exceeded|model_not_found|rate_limit_error|provider_error|timeout_error)) (required) ("invalid_request_error"|"authentication_error"|"budget_exceeded"|"model_not_found"|"rate_limit_error"|"provider_error"|"timeout_error")
  - **code** (integer)

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "anthropic/claude-opus-4.8",
  "baseline_model": "anthropic/claude-opus-4.8",
  "messages": [
    {
      "role": "system",
      "content": "value"
    }
  ],
  "stream": "false",
  "temperature": "0",
  "max_tokens": "0",
  "tools": "item1,item2",
  "tool_choice": "value",
  "response_format": "value",
  "webhook_url": "string"
}'
```

```

--------------------------------

### POST /batches

Source: https://aigateway.sh/openapi.json

Submits a JSONL file (uploaded via /files with purpose=batch) for batched inference at 50% off. SLA 24h.

```markdown
### Request Body

**Content-Type:** application/json

- **input_file_id** (string) (required)
- **endpoint** (string) (required) (example: "/v1/chat/completions")

### Responses

#### 200 - Batch record

**Batch**
- **id** (string)
- **object** (string (batch)) ("batch")
- **endpoint** (string)
- **input_file_id** (string)
- **completion_window** (string) (example: "24h")
- **status** (string (validating|queued|in_progress|finalizing|completed|failed|expired)) ("validating"|"queued"|"in_progress"|"finalizing"|"completed"|"failed"|"expired")
- **output_file_id** (string)
- **error_file_id** (string)
- **request_counts** (object)
  - **processing** (integer)
  - **completed** (integer)
  - **failed** (integer)
- **created_at** (integer)
- **expires_at** (integer)
- **completed_at** (integer)

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/batches" \
  -H "Content-Type: application/json" \
  -d '{
  "input_file_id": "string",
  "endpoint": "/v1/chat/completions"
}'
```

```

--------------------------------

### Schema: File

Source: https://aigateway.sh/openapi.json

Schema definition for File

```markdown
## Schema: File

Schema definition for File

**Type:** object

- **id** (string) (example: "file-abc")
- **object** (string (file)) ("file")
- **bytes** (integer)
- **created_at** (integer)
- **filename** (string)
- **purpose** (string)

```

--------------------------------

### POST /sub-accounts

Source: https://aigateway.sh/openapi.json

Mints a scoped key for one of your end customers. Spend cap, RPM, default tag, isolated analytics.

```markdown
### Request Body

**Content-Type:** application/json

- **name** (string) (required)
- **external_ref** (string)
- **spend_cap_cents** (integer) (required)
- **rate_limit_rpm** (integer)
- **default_tag** (string)

### Responses

#### 200 - Sub-account record (key shown once)

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/sub-accounts" \
  -H "Content-Type: application/json" \
  -d '{
  "name": "string",
  "external_ref": "string",
  "spend_cap_cents": "0",
  "rate_limit_rpm": "0",
  "default_tag": "string"
}'
```

```

--------------------------------

### POST /audio/music

Source: https://aigateway.sh/openapi.json

API endpoint for POST /audio/music

```markdown
### Request Body

**Content-Type:** application/json

- **prompt** (string) (required)
- **model** (string)
- **duration** (number)
- **aspect_ratio** (string)
- **resolution** (string)
- **image_url** (string (uri))
- **webhook_url** (string (uri))

### Responses

#### 202 - Job accepted

**JobAccepted**
- **id** (string) (required) (example: "job_abc123")
- **status** (string (queued)) (required) ("queued")
- **object** (string (job)) (required) ("job")

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/audio/music" \
  -H "Content-Type: application/json" \
  -d '{
  "prompt": "string",
  "model": "string",
  "duration": "0",
  "aspect_ratio": "string",
  "resolution": "string",
  "image_url": "string",
  "webhook_url": "string"
}'
```

```

--------------------------------

### Schema: Batch

Source: https://aigateway.sh/openapi.json

Schema definition for Batch

```markdown
## Schema: Batch

Schema definition for Batch

**Type:** object

- **id** (string)
- **object** (string (batch)) ("batch")
- **endpoint** (string)
- **input_file_id** (string)
- **completion_window** (string) (example: "24h")
- **status** (string (validating|queued|in_progress|finalizing|completed|failed|expired)) ("validating"|"queued"|"in_progress"|"finalizing"|"completed"|"failed"|"expired")
- **output_file_id** (string)
- **error_file_id** (string)
- **request_counts** (object)
  - **processing** (integer)
  - **completed** (integer)
  - **failed** (integer)
- **created_at** (integer)
- **expires_at** (integer)
- **completed_at** (integer)

```

--------------------------------

### POST /files

Source: https://aigateway.sh/openapi.json

API endpoint for POST /files

```markdown
### Request Body

**Content-Type:** multipart/form-data

- **file** (string (binary)) (required)
- **purpose** (string (batch|batch_output|fine-tune|assistants)) (required) ("batch"|"batch_output"|"fine-tune"|"assistants")

### Responses

#### 200 - File metadata

**File**
- **id** (string) (example: "file-abc")
- **object** (string (file)) ("file")
- **bytes** (integer)
- **created_at** (integer)
- **filename** (string)
- **purpose** (string)

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/files" \
  -H "Content-Type: application/json" \
  -d '{
  "file": "string",
  "purpose": "batch"
}'
```

```

--------------------------------

### POST /replays

Source: https://aigateway.sh/openapi.json

API endpoint for POST /replays

```markdown
### Request Body

**Content-Type:** application/json

- **source_request_id** (string) (required)
- **target_model** (string) (required)
- **shadow** (boolean)

### Responses

#### 200 - Side-by-side cost / latency / output diff

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/replays" \
  -H "Content-Type: application/json" \
  -d '{
  "source_request_id": "string",
  "target_model": "string",
  "shadow": "false"
}'
```

```

--------------------------------

### API Overview: AIgateway

Source: https://aigateway.sh/openapi.json

OpenAI-compatible endpoints for chat, embeddings, images, audio, video, translation, classification, OCR, rerank, moderation, plus aggregator-native primitives (sub-accounts, evals, replays, batches, files, signed URLs, webhooks). Catalog: 1000+ models across 85+ labs. Change `base_url` on the OpenAI SDK to https://api.aigateway.sh/v1 and every existing integration works.

```yaml
# AIgateway
# Version: 1

OpenAI-compatible endpoints for chat, embeddings, images, audio, video, translation, classification, OCR, rerank, moderation, plus aggregator-native primitives (sub-accounts, evals, replays, batches, files, signed URLs, webhooks). Catalog: 1000+ models across 85+ labs. Change `base_url` on the OpenAI SDK to https://api.aigateway.sh/v1 and every existing integration works.

# Base URL: https://api.aigateway.sh/v1
```

--------------------------------

### Schema: AsyncJobRequest

Source: https://aigateway.sh/openapi.json

Schema definition for AsyncJobRequest

```markdown
## Schema: AsyncJobRequest

Schema definition for AsyncJobRequest

**Type:** object

- **prompt** (string) (required)
- **model** (string)
- **duration** (number)
- **aspect_ratio** (string)
- **resolution** (string)
- **image_url** (string (uri))
- **webhook_url** (string (uri))

```

--------------------------------

### POST /evals

Source: https://aigateway.sh/openapi.json

Run an eval across candidate models on your dataset. Returns an id you can use as `model: 'eval:<id>'` to always route to the current winner.

```markdown
### Request Body

**Content-Type:** application/json

- **name** (string) (required)
- **candidate_models** (array (string)) (required)
- **dataset** (array (object)) (required)
- **metric** (string (quality|cost|speed)) (required) ("quality"|"cost"|"speed")

### Responses

#### 200 - Eval run started

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/evals" \
  -H "Content-Type: application/json" \
  -d '{
  "name": "string",
  "candidate_models": [
    "string"
  ],
  "dataset": [
    "value"
  ],
  "metric": "quality"
}'
```

```

--------------------------------

### POST /audio/speech

Source: https://aigateway.sh/openapi.json

API endpoint for POST /audio/speech

```markdown
### Request Body

**Content-Type:** application/json

- **model** (string) (required) (example: "deepgram/aura-2-en")
- **input** (string) (required)
- **voice** (string)
- **response_format** (string (mp3|opus|aac|flac)) ("mp3"|"opus"|"aac"|"flac")

### Responses

#### 200 - Audio binary stream

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/audio/speech" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepgram/aura-2-en",
  "input": "string",
  "voice": "string",
  "response_format": "mp3"
}'
```

```

--------------------------------

### POST /detections

Source: https://aigateway.sh/openapi.json

API endpoint for POST /detections

```markdown
### Request Body

**Content-Type:** application/json

- **image_url** (string (uri))
- **image_b64** (string)
- **model** (string) (example: "facebook/detr-resnet-50")

### Responses

#### 200 - Bounding boxes + labels

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/detections" \
  -H "Content-Type: application/json" \
  -d '{
  "image_url": "string",
  "image_b64": "string",
  "model": "facebook/detr-resnet-50"
}'
```

```

--------------------------------

### Schema: WebhookEvent

Source: https://aigateway.sh/openapi.json

Schema definition for WebhookEvent

```markdown
## Schema: WebhookEvent

Schema definition for WebhookEvent

**Type:** object

- **id** (string) (required) (example: "evt_01jx...")
- **type** (string (job.completed|job.failed|balance.low|balance.exhausted|usage.threshold.exceeded|usage.daily.summary|subaccount.created|subaccount.spend.exceeded|model.added|model.deprecated|key.rotated)) (required) ("job.completed"|"job.failed"|"balance.low"|"balance.exhausted"|"usage.threshold.exceeded"|"usage.daily.summary"|"subaccount.created"|"subaccount.spend.exceeded"|"model.added"|"model.deprecated"|"key.rotated")
- **occurred_at** (integer) (required): Unix seconds
- **data** (object) (required): Event-type-specific payload

```

--------------------------------

### POST /budgets

Source: https://aigateway.sh/openapi.json

API endpoint for POST /budgets

```markdown
### Request Body

**Content-Type:** application/json

- **tag** (string) (required)
- **cap_cents** (integer) (required)

### Responses

#### 200 - Budget created

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/budgets" \
  -H "Content-Type: application/json" \
  -d '{
  "tag": "string",
  "cap_cents": "0"
}'
```

```

--------------------------------

### Schema: Job

Source: https://aigateway.sh/openapi.json

Schema definition for Job

```markdown
## Schema: Job

Schema definition for Job

**Type:** object

- **id** (string) (required)
- **status** (string (queued|processing|completed|failed)) (required) ("queued"|"processing"|"completed"|"failed")
- **modality** (string)
- **model** (string)
- **created_at** (integer)
- **updated_at** (integer)
- **error** (object)
  - **message** (string)
  - **code** (string)
- **result_file_id** (string)
- **result_url** (string (uri))
- **webhook_url** (string (uri))
- **webhook_delivered** (boolean)
- **attempts** (integer)

```

--------------------------------

### DELETE /cache

Source: https://aigateway.sh/openapi.json

Drops every entry in the exact-match KV cache and the semantic Vectorize cache for the authenticated account.

```markdown
### Responses

#### 200 - Purged

Empty response body

### Example Usage

```bash
curl -X DELETE "https://api.aigateway.sh/v1/cache"
```

```

--------------------------------

### POST /3d/generations

Source: https://aigateway.sh/openapi.json

API endpoint for POST /3d/generations

```markdown
### Request Body

**Content-Type:** application/json

- **prompt** (string) (required)
- **model** (string)
- **duration** (number)
- **aspect_ratio** (string)
- **resolution** (string)
- **image_url** (string (uri))
- **webhook_url** (string (uri))

### Responses

#### 202 - Job accepted

**JobAccepted**
- **id** (string) (required) (example: "job_abc123")
- **status** (string (queued)) (required) ("queued")
- **object** (string (job)) (required) ("job")

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/3d/generations" \
  -H "Content-Type: application/json" \
  -d '{
  "prompt": "string",
  "model": "string",
  "duration": "0",
  "aspect_ratio": "string",
  "resolution": "string",
  "image_url": "string",
  "webhook_url": "string"
}'
```

```

--------------------------------

### Schema: JobAccepted

Source: https://aigateway.sh/openapi.json

Schema definition for JobAccepted

```markdown
## Schema: JobAccepted

Schema definition for JobAccepted

**Type:** object

- **id** (string) (required) (example: "job_abc123")
- **status** (string (queued)) (required) ("queued")
- **object** (string (job)) (required) ("job")

```

--------------------------------

### POST /moderations

Source: https://aigateway.sh/openapi.json

API endpoint for POST /moderations

```markdown
### Request Body

**Content-Type:** application/json

- **model** (string) (example: "meta/llama-guard-3-8b")
- **input** (string) (required)

### Responses

#### 200 - Per-category safety flags

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/moderations" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "meta/llama-guard-3-8b",
  "input": "value"
}'
```

```

--------------------------------

### DELETE /files/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for DELETE /files/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Deleted

Empty response body

### Example Usage

```bash
curl -X DELETE "https://api.aigateway.sh/v1/files/{id}"
```

```

--------------------------------

### Schema: ChatCompletionRequest

Source: https://aigateway.sh/openapi.json

Schema definition for ChatCompletionRequest

```markdown
## Schema: ChatCompletionRequest

Schema definition for ChatCompletionRequest

**Type:** object

- **model** (string) (required): Provider-qualified slug, or "auto" / "auto/<modality>" to let the Auto Router pick the cheapest model in a curated pool that clears the quality floor (never above your baseline). Featured: anthropic/claude-opus-4.8, minimax/m3, google/gemini-3.1-pro, anthropic/claude-sonnet-4.6, openai/gpt-5.5, google/gemini-2.5-pro, moonshot/kimi-k2.6, xai/grok-4.20-multi-agent-0309, openai/gpt-5.4-mini, alibaba/qwen3.5-397b-a17b. Fetch full list from /models. Catalog: 1000+. (example: "anthropic/claude-opus-4.8")
- **baseline_model** (string): Auto Router only: the model you'd otherwise call. The router routes DOWN from it, so it doubles as a hard cost ceiling — you never pay more than calling it directly. Defaults to the premium model for the modality. Used only when model is "auto" / "auto/<modality>" or omitted. (example: "anthropic/claude-opus-4.8")
- **messages** (array (object)) (required)
  Array items:
    - **role** (string (system|user|assistant|tool)) (required) ("system"|"user"|"assistant"|"tool")
    - **content** (string) (required)
- **stream** (boolean)
- **temperature** (number)
- **max_tokens** (integer)
- **tools** (array)
- **tool_choice** (string)
- **response_format** (object)
- **webhook_url** (string (uri)): If set, delivery flips to async-job mode (202 + signed callback)

```

--------------------------------

### POST /embeddings

Source: https://aigateway.sh/openapi.json

API endpoint for POST /embeddings

```markdown
### Request Body

**Content-Type:** application/json

- **model** (string) (required) (example: "baai/bge-m3")
- **input** (string) (required)

### Responses

#### 200 - Embedding vectors

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/embeddings" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "baai/bge-m3",
  "input": "value"
}'
```

```

--------------------------------

### POST /rerank

Source: https://aigateway.sh/openapi.json

API endpoint for POST /rerank

```markdown
### Request Body

**Content-Type:** application/json

- **query** (string) (required)
- **documents** (array (string)) (required)
- **model** (string) (example: "baai/bge-reranker-base")

### Responses

#### 200 - Documents ordered by relevance

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/rerank" \
  -H "Content-Type: application/json" \
  -d '{
  "query": "string",
  "documents": [
    "string"
  ],
  "model": "baai/bge-reranker-base"
}'
```

```

--------------------------------

### POST /images/generations

Source: https://aigateway.sh/openapi.json

API endpoint for POST /images/generations

```markdown
### Request Body

**Content-Type:** application/json

- **model** (string) (required) (example: "black-forest-labs/flux-2-klein-9b")
- **prompt** (string) (required)
- **size** (string) (example: "1024x1024")
- **quality** (string (standard|hd)) ("standard"|"hd")
- **style** (string)
- **response_format** (string (url|b64_json)) ("url"|"b64_json")
- **n** (integer)

### Responses

#### 200 - Image URL(s) or base64

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/images/generations" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "black-forest-labs/flux-2-klein-9b",
  "prompt": "string",
  "size": "1024x1024",
  "quality": "standard",
  "style": "string",
  "response_format": "url",
  "n": "1"
}'
```

```

--------------------------------

### POST /webhook-secret/rotate

Source: https://aigateway.sh/openapi.json

API endpoint for POST /webhook-secret/rotate

```markdown
### Responses

#### 200 - New secret

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/webhook-secret/rotate"
```

```

--------------------------------

### POST /batches/{id}/cancel

Source: https://aigateway.sh/openapi.json

API endpoint for POST /batches/{id}/cancel

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Cancelled

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/batches/{id}/cancel"
```

```

--------------------------------

### Security: bearerAuth

Source: https://aigateway.sh/openapi.json

Security scheme: bearerAuth

```markdown
## Security: bearerAuth

**Description:** Security scheme: bearerAuth

**Type:** http

**Scheme:** bearer

**Bearer Format:** sk-aig-...


```

--------------------------------

### POST /translations

Source: https://aigateway.sh/openapi.json

API endpoint for POST /translations

```markdown
### Request Body

**Content-Type:** application/json

- **text** (string) (required)
- **source_lang** (string) (required) (example: "en")
- **target_lang** (string) (required) (example: "fr")
- **model** (string) (example: "meta/m2m100-1.2b")

### Responses

#### 200 - Translated text

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/translations" \
  -H "Content-Type: application/json" \
  -d '{
  "text": "string",
  "source_lang": "en",
  "target_lang": "fr",
  "model": "meta/m2m100-1.2b"
}'
```

```

--------------------------------

### POST /ocr

Source: https://aigateway.sh/openapi.json

API endpoint for POST /ocr

```markdown
### Request Body

**Content-Type:** application/json

- **image_url** (string (uri))
- **image_b64** (string)
- **model** (string) (example: "microsoft/trocr-base-printed")

### Responses

#### 200 - Extracted text

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/ocr" \
  -H "Content-Type: application/json" \
  -d '{
  "image_url": "string",
  "image_b64": "string",
  "model": "microsoft/trocr-base-printed"
}'
```

```

--------------------------------

### PATCH /sub-accounts/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for PATCH /sub-accounts/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Updated record

Empty response body

### Example Usage

```bash
curl -X PATCH "https://api.aigateway.sh/v1/sub-accounts/{id}"
```

```

--------------------------------

### DELETE /jobs/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for DELETE /jobs/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Cancelled

Empty response body

### Example Usage

```bash
curl -X DELETE "https://api.aigateway.sh/v1/jobs/{id}"
```

```

--------------------------------

### POST /classifications

Source: https://aigateway.sh/openapi.json

API endpoint for POST /classifications

```markdown
### Request Body

**Content-Type:** application/json

- **input** (string) (required)
- **model** (string) (example: "huggingface/distilbert-sst-2")

### Responses

#### 200 - Label + score per input

Empty response body

### Example Usage

```bash
curl -X POST "https://api.aigateway.sh/v1/classifications" \
  -H "Content-Type: application/json" \
  -d '{
  "input": "value",
  "model": "huggingface/distilbert-sst-2"
}'
```

```

--------------------------------

### DELETE /sub-accounts/{id}

Source: https://aigateway.sh/openapi.json

API endpoint for DELETE /sub-accounts/{id}

```markdown
### Parameters

- **id** (string, path, required)

### Responses

#### 200 - Deleted

Empty response body

### Example Usage

```bash
curl -X DELETE "https://api.aigateway.sh/v1/sub-accounts/{id}"
```

```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.