Eden AI V3

https://www.edenai.co/docs/v3
Admin
Eden AI V3 is a next-generation unified API that simplifies AI integration with OpenAI-compatible...
Tokens:14,020
Snippets:68
Trust Score:9.8
Update:1 week ago
Show doc for...
Tokens:
Raw
### Use the OpenAI SDK with Eden AI

Source: https://www.edenai.co/docs/v3/quickstart/first-llm-call

Shows how to use the official OpenAI SDK with Eden AI by setting the base_url to https://api.edenai.run/v3. Replace YOUR_EDEN_AI_API_KEY with your actual key. The example uses the vertex/gemini-flash-latest model.

```python
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_EDEN_AI_API_KEY",
    base_url="https://api.edenai.run/v3"
)

response = client.chat.completions.create(
    model="vertex/gemini-flash-latest",
    messages=[{"role": "user", "content": "Hello!"}]
)

print(response.choices[0].message.content)

```

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

### POST /v3/universal-ai/async

Source: https://www.edenai.co/docs/v3/expert-models/features/audio/speech-to-text-async

Submits an audio file for asynchronous speech-to-text transcription. Returns a job ID that can be polled with GET /v3/universal-ai/async/{job_id}.

```APIDOC
## POST /v3/universal-ai/async

### Description
Submits an audio file for asynchronous speech-to-text transcription. The initial response returns a job ID; poll `GET /v3/universal-ai/async/{job_id}` until the job completes.

### Method
POST

### Endpoint
`/v3/universal-ai/async`

### Parameters
#### Request Body
- **model** (string) - Required - Model string pattern: `audio/speech_to_text_async/{provider}[/{model}]`
- **input** (object) - Required - Contains the following fields:
  - **file** (file_input) - Required - Audio file ID from /v3/upload or direct file URL
  - **language** (string) - Optional - Language code in ISO format (e.g., 'en', 'fr', 'es')
  - **speakers** (int) - Optional - Number of speakers present in the audio
  - **profanity_filter** (bool) - Optional - Whether to filter profanity and replace inappropriate words with a series of asterisks
  - **vocabulary** (array[string]) - Optional - List of words or composed words to be detected by the speech to text engine

### Request Example
```json
{
  "model": "audio/speech_to_text_async/amazon",
  "input": {
    "file": "YOUR_FILE_UUID_OR_URL",
    "language": "en"
  }
}
```

### Response
#### Success Response (200)
- **text** (string) - Required - The transcribed text
- **diarization** (object) - Required - Contains speaker diarization information
  - **total_speakers** (int) - Required - Total number of speakers detected
  - **entries** (array[object]) - Optional - List of diarization entries
    - **segment** (string) - Required - The transcribed segment
    - **start_time** (string) - Required - Start time of the segment
    - **end_time** (string) - Required - End time of the segment
    - **speaker** (int) - Required - Speaker identifier
    - **confidence** (float) - Required - Confidence score
- **error_message** (string) - Optional - Error message if any

#### Response Example
```json
{
  "text": "...",
  "diarization": {
    "total_speakers": 2,
    "entries": [
      {
        "segment": "...",
        "start_time": "...",
        "end_time": "...",
        "speaker": 1,
        "confidence": 0.95
      }
    ]
  }
}
```

### Available Providers
| Provider | Model String | Price |
|---|---|---|
| amazon | `audio/speech_to_text_async/amazon` | $0.024 per 60 seconds |
| assembly | `audio/speech_to_text_async/assembly` | $0.011 per 60 seconds |
| deepgram (base) | `audio/speech_to_text_async/deepgram/base` | $0.0169 per 60 seconds |
| deepgram | `audio/speech_to_text_async/deepgram` | $0.0189 per 60 seconds |
| deepgram (enhanced) | `audio/speech_to_text_async/deepgram/enhanced` | $0.0189 per 60 seconds |
| deepgram (nova-3) | `audio/speech_to_text_async/deepgram/nova-3` | $0.0052 per 60 seconds |
| gladia | `audio/speech_to_text_async/gladia` | $0.0102 per 60 seconds |
| google | `audio/speech_to_text_async/google` | $0.024 per 60 seconds |
| microsoft | `audio/speech_to_text_async/microsoft` | $0.0168 per 60 seconds |
| openai | `audio/speech_to_text_async/openai` | $0.006 per 60 seconds |

### Quick Start
> This is an **async** feature. The initial response returns a job ID. Poll `GET /v3/universal-ai/async/{job_id}` until the job completes.

#### Python
```python
import requests

url = "https://api.edenai.run/v3/universal-ai/async"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "audio/speech_to_text_async/amazon",
    "input": {
        "file": "YOUR_FILE_UUID_OR_URL",
        "language": "en"
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())
```

#### cURL
```bash
curl -X POST https://api.edenai.run/v3/universal-ai/async \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "audio/speech_to_text_async/amazon",
    "input": {"file": "YOUR_FILE_UUID_OR_URL", "language": "en"}
  }'
```
```

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

### POST /v3/universal-ai/async

Source: https://www.edenai.co/docs/v3/expert-models/features/video/deepfake-detection-async

Submits a video file for asynchronous deepfake detection. Returns a job ID that can be polled using GET /v3/universal-ai/async/{job_id}.

```APIDOC
## POST /v3/universal-ai/async

### Description
Submits a video file for asynchronous deepfake detection. Returns a job ID that can be polled using GET /v3/universal-ai/async/{job_id}.

### Method
POST

### Endpoint
/v3/universal-ai/async

### Parameters
#### Request Body
- **model** (string) - Required - Model string pattern: `video/deepfake_detection_async/{provider}[/{model}]`
- **input.file** (file_input) - Required - Video file ID from /v3/upload or direct file URL

### Request Example
{
  "model": "video/deepfake_detection_async/resemble",
  "input": {
    "file": "YOUR_FILE_UUID_OR_URL"
  }
}

### Response
#### Success Response (200)
- **average_score** (float) - Required - Average deepfake score
- **prediction** (enum) - Required - Prediction result
- **details_per_frame** (array[object]) - Optional - Per-frame details
  - **position** (float) - Required - Frame position
  - **score** (float) - Required - Frame score
  - **prediction** (enum) - Required - Frame prediction

#### Response Example
{
  "job_id": "string"
}
```

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

### POST /v3/universal-ai/async

Source: https://www.edenai.co/docs/v3/expert-models/features/video/generation-async

Submits an asynchronous video generation job. The initial response returns a job ID, which can be polled using the GET /v3/universal-ai/async/{job_id} endpoint until the job completes.

```APIDOC
## POST /v3/universal-ai/async

### Description
Submits an asynchronous video generation job. The initial response returns a job ID, which can be polled using the GET /v3/universal-ai/async/{job_id} endpoint until the job completes.

### Method
POST

### Endpoint
https://api.edenai.run/v3/universal-ai/async

### Parameters
#### Request Body
- **model** (string) - Required - The model identifier for video generation, e.g., "video/generation_async/amazon/amazon.nova-reel-v1:0".
- **input** (object) - Required - The input parameters for the generation.
  - **text** (string) - Required - The text prompt describing the video to generate.

### Request Example
{
  "model": "video/generation_async/amazon/amazon.nova-reel-v1:0",
  "input": {
    "text": "The quick brown fox jumps over the lazy dog."
  }
}

### Response
#### Success Response (200)
- **job_id** (string) - The job ID for polling the async job status.

#### Response Example
{
  "job_id": "your-job-id"
}
```

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

### Example chat completion response

Source: https://www.edenai.co/docs/v3/quickstart/first-llm-call

Shows the JSON structure returned by the chat completions endpoint, including the generated message, finish reason, and token usage.

```json
{
  "id": "chatcmpl-123",
  "object": "chat.completion",
  "choices": [
    {
      "message": {"role": "assistant", "content": "Hello! I'm doing well, thank you."},
      "index": 0,
      "finish_reason": "stop"
    }
  ],
  "usage": {"prompt_tokens": 12, "completion_tokens": 10, "total_tokens": 22}
}

```

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

### Response: Universal AI call result

Source: https://www.edenai.co/docs/v3/quickstart/first-expert-model-call

Example JSON response from the Universal AI endpoint for a text moderation call. Shows the status, cost, provider, feature, subfeature, and output with moderation details.

```JSON
{
  "status": "success",
  "cost": 0.0001,
  "provider": "google",
  "feature": "text",
  "subfeature": "moderation",
  "output": {
    "nsfw_likelihood": 1,
    "items": [
      {"label": "Toxic", "likelihood": 1, "likelihood_score": 0.0908}
    ],
    "nsfw_likelihood_score": 0.0908
  }
}

```

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

### POST /v3/universal-ai/async

Source: https://www.edenai.co/docs/v3/expert-models/features/video/generation-async

Initiates an asynchronous video generation job. Provide a text prompt and optional parameters to generate a video using the specified provider and model.

```APIDOC
## POST /v3/universal-ai/async

### Description
Initiates an asynchronous video generation job. Provide a text prompt and optional parameters to generate a video using the specified provider and model.

### Method
POST

### Endpoint
/v3/universal-ai/async

### Parameters
#### Request Body
- **text** (string) - Required - Text prompt describing the content and style of the video to generate
- **file** (file_input) - Optional - Initial keyframe image (JPEG or PNG) file ID from /v3/upload or direct file URL
- **duration** (int) - Optional - Length of the generated video in seconds
- **dimension** (string) - Optional - Video resolution in 'widthxheight' format (e.g., '1280x720')
- **seed** (int) - Optional - Random seed for generation reproducibility (0 to 2,147,483,646). Omit for varied output.

### Request Example
{
  "text": "A serene landscape with mountains and a lake",
  "duration": 5,
  "dimension": "1280x720"
}

### Response
#### Success Response (200)
- **video** (string) - The generated video data or identifier
- **video_resource_url** (string) - URL to access the generated video resource
- **usage** (dict) - Optional - Usage information

#### Response Example
{
  "video": "video_12345",
  "video_resource_url": "https://example.com/videos/video_12345.mp4",
  "usage": {
    "seconds": 5
  }
}
```

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

### Python: Generate image with requests

Source: https://www.edenai.co/docs/v3/expert-models/features/image/generation

Sends a POST request to the Eden AI universal endpoint with the image generation model and input text. Requires an API key and the requests library.

```python
import requests

url = "https://api.edenai.run/v3/universal-ai"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "image/generation/bytedance",
    "input": {
        "text": "The quick brown fox jumps over the lazy dog."
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

```

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

### Submit async video generation job with Python

Source: https://www.edenai.co/docs/v3/expert-models/features/video/generation-async

Sends a POST request to the async endpoint with the video generation model and input text. The response contains a job ID that must be polled until completion.

```python
import requests

url = "https://api.edenai.run/v3/universal-ai/async"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "video/generation_async/amazon/amazon.nova-reel-v1:0",
    "input": {
        "text": "The quick brown fox jumps over the lazy dog."
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

```

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

### Python: Text-to-Speech with requests

Source: https://www.edenai.co/docs/v3/expert-models/features/audio/tts

Sends a POST request to the Eden AI Universal AI endpoint to synthesize speech from text using the amazon/neural model. Requires an API key and the requests library. Prints the JSON response.

```python
import requests

url = "https://api.edenai.run/v3/universal-ai"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "audio/tts/amazon/neural",
    "input": {
        "text": "The quick brown fox jumps over the lazy dog."
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

```

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

### POST /v3/universal-ai

Source: https://www.edenai.co/docs/v3/expert-models/features/image/generation

Generates images from a text prompt using a specified provider and model. The model is specified via the `model` string pattern `image/generation/{provider}[/{model}]`.

```APIDOC
## POST /v3/universal-ai

### Description
Generates images from a text prompt using a specified provider and model. The model is specified via the `model` string pattern `image/generation/{provider}[/{model}]`.

### Method
POST

### Endpoint
`/v3/universal-ai`

### Parameters
#### Request Body
- **text** (string) - Required - Text prompt for image generation
- **resolution** (string) - Optional - Image resolution as a named size (e.g. '1024x1024'). Optional: providers may also accept native sizing knobs (width/height, aspect_ratio, image_size tier, etc.) via `provider_params`, in which case `provider_params` takes precedence.
- **num_images** (int) - Optional - Number of images to generate
- **reference_images** (array[file_input | object]) - Optional - Optional reference images. Plain file IDs/URLs are treated as ReferenceImage with kind=None. Support and per-model limits vary by model; models that do not accept reference images surface the error as status="fail" with a provider message in the response body.
- **mask** (file_input) - Optional - Optional inpainting mask. Must be paired with at least one reference image (otherwise the request returns 422). Support varies by model; models that do not accept a mask surface the error as status="fail" with a provider message in the response body.

### Response
#### Success Response (200)
- **image** (string) - The generated image (likely a URL or base64 string)
- **image_resource_url** (string) - URL to the image resource

#### Response Example
```json
{
  "image": "string",
  "image_resource_url": "string"
}
```
```

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

### Model Format: feature/subfeature/provider[/model]

Source: https://www.edenai.co/docs/v3/quickstart/first-expert-model-call

The model format for the Universal AI endpoint. Use the pattern feature/subfeature/provider[/model] to specify a model.

```text
feature/subfeature/provider[/model]

```

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

### Python: Submit deepfake detection job

Source: https://www.edenai.co/docs/v3/expert-models/features/video/deepfake-detection-async

Sends a POST request to the async endpoint with the model 'video/deepfake_detection_async/resemble' and a file UUID or URL. The response contains a job ID that must be polled until completion.

```python
import requests

url = "https://api.edenai.run/v3/universal-ai/async"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "video/deepfake_detection_async/resemble",
    "input": {
        "file": "YOUR_FILE_UUID_OR_URL"
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

```

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

### Make a chat completion request with JavaScript

Source: https://www.edenai.co/docs/v3/quickstart/first-llm-call

Sends a chat completion request to Eden AI's OpenAI-compatible endpoint using the fetch API. Replace YOUR_API_KEY with your actual token. The response is parsed as JSON and the assistant's message content is logged to the console.

```javascript
const response = await fetch('https://api.edenai.run/v3/chat/completions', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    model: 'openai/gpt-latest',
    messages: [{role: 'user', content: 'Hello!'}]
  })
});

const result = await response.json();
console.log(result.choices[0].message.content);

```

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

### Submit async video generation job with cURL

Source: https://www.edenai.co/docs/v3/expert-models/features/video/generation-async

Sends a POST request to the async endpoint using cURL with the same model and input text. The response contains a job ID that must be polled until completion.

```bash
curl -X POST https://api.edenai.run/v3/universal-ai/async \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "video/generation_async/amazon/amazon.nova-reel-v1:0",
    "input": {"text": "The quick brown fox jumps over the lazy dog."}
  }'

```

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

### POST /v3/universal-ai

Source: https://www.edenai.co/docs/v3/expert-models/features/audio/tts

Converts text to speech using the specified provider and model. The model string follows the pattern audio/tts/{provider}[/{model}]. Returns a URL to the generated audio resource.

```APIDOC
## POST /v3/universal-ai

### Description
Converts text to speech using the specified provider and model. The model string follows the pattern `audio/tts/{provider}[/{model}]`. Returns a URL to the generated audio resource.

### Method
POST

### Endpoint
`/v3/universal-ai`

### Parameters
#### Request Body
- **model** (string) - Required - Model string pattern: `audio/tts/{provider}[/{model}]`
- **input** (object) - Required - Input object containing the following fields:
  - **text** (string) - Required - Text to convert to speech
  - **voice** (string) - Optional - Voice identifier to use for synthesis, can be name or id. For supported voices for your provider/model, please check each provider’s documentation.
  - **speed** (float) - Optional - Speed multiplier for speech synthesis
  - **audio_format** (string) - Optional - Desired audio format (e.g., 'mp3', 'wav')
  - **speaking_pitch** (int) - Optional - Increase or decrease the speaking pitch by a percentage from -100% to 100%, 0 is normal.
  - **speaking_volume** (int) - Optional - Increase or decrease the audio volume by a percentage from -100% to 100%, 0 is normal.

### Request Example
```json
{
  "model": "audio/tts/amazon/neural",
  "input": {
    "text": "The quick brown fox jumps over the lazy dog."
  }
}
```

### Response
#### Success Response (200)
- **audio_resource_url** (string) - URL to the generated audio resource

#### Response Example
```json
{
  "audio_resource_url": "https://example.com/audio.mp3"
}
```

### Available Providers
| Provider | Model String | Price |
|---|---|---|
| amazon (neural) | `audio/tts/amazon/neural` | $0.016 per 1,000 chars |
| amazon (standard) | `audio/tts/amazon/standard` | $0.004 per 1,000 chars |
| deepgram (aura) | `audio/tts/deepgram/aura` | $0.015 per 1,000 chars |
| deepgram (aura-2) | `audio/tts/deepgram/aura-2` | $0.03 per 1,000 chars |
| elevenlabs | `audio/tts/elevenlabs` | $0.3 per 1,000 chars |
| elevenlabs (eleven_flash_v2) | `audio/tts/elevenlabs/eleven_flash_v2` | $0.3 per 1,000 chars |
| elevenlabs (eleven_flash_v2_5) | `audio/tts/elevenlabs/eleven_flash_v2_5` | $0.3 per 1,000 chars |
| elevenlabs (eleven_multilingual_v2) | `audio/tts/elevenlabs/eleven_multilingual_v2` | $0.3 per 1,000 chars |
| elevenlabs (eleven_turbo_v2) | `audio/tts/elevenlabs/eleven_turbo_v2` | $0.3 per 1,000 chars |
| elevenlabs (eleven_turbo_v2_5) | `audio/tts/elevenlabs/eleven_turbo_v2_5` | $0.3 per 1,000 chars |
| elevenlabs (eleven_v3) | `audio/tts/elevenlabs/eleven_v3` | $0.3 per 1,000 chars |
| google (casual) | `audio/tts/google/casual` | $16 per 1,000,000 chars |
| google (chirp3-hd) | `audio/tts/google/chirp3-hd` | $30 per 1,000,000 chars |
| google (chirp-hd) | `audio/tts/google/chirp-hd` | $30 per 1,000,000 chars |
| google (gemini-2.5-flash-tts) | `audio/tts/google/gemini-2.5-flash-tts` | $0.006 per minute |
| google (gemini-2.5-pro-tts) | `audio/tts/google/gemini-2.5-pro-tts` | $0.012 per minute |
| google (gemini-3.1-flash-tts-preview) | `audio/tts/google/gemini-3.1-flash-tts-preview` | $0.012 per minute |
| google (neural2) | `audio/tts/google/neural2` | $16 per 1,000,000 chars |
| google (news) | `audio/tts/google/news` | $160 per 1,000,000 chars |
| google (polyglot) | `audio/tts/google/polyglot` | $16 per 1,000,000 chars |
| google (standard) | `audio/tts/google/standard` | $4 per 1,000,000 chars |
| google (studio) | `audio/tts/google/studio` | $160 per 1,000,000 chars |
| google (wavenet) | `audio/tts/google/wavenet` | $4 per 1,000,000 chars |
| lovoai | `audio/tts/lovoai` | $160 per 1,000,000 chars |
| microsoft (neural) | `audio/tts/microsoft/neural` | $16 per 1,000,000 chars |
| microsoft (neural-hd) | `audio/tts/microsoft/neural-hd` | $30 per 1,000,000 chars |
| openai (gpt-4o-mini-tts) | `audio/tts/openai/gpt-4o-mini-tts` | $0.015 per minute |
| openai (tts-1) | `audio/tts/openai/tts-1` | $15 per 1,000,000 chars |
| openai (tts-1-hd) | `audio/tts/openai/tts-1-hd` | $30 per 1,000,000 chars |

### Quick Start
#### Python
```python
import requests

url = "https://api.edenai.run/v3/universal-ai"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "audio/tts/amazon/neural",
    "input": {
        "text": "The quick brown fox jumps over the lazy dog."
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())
```

#### cURL
```bash
curl -X POST https://api.edenai.run/v3/universal-ai \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "audio/tts/amazon/neural",
    "input": {"text": "The quick brown fox jumps over the lazy dog."}
  }'
```
```

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

### Make a chat completion request with Python

Source: https://www.edenai.co/docs/v3/quickstart/first-llm-call

Sends a chat completion request to Eden AI's OpenAI-compatible endpoint using the requests library. Replace YOUR_API_KEY with your actual token. The response is parsed as JSON and the assistant's message content is printed.

```python
import requests

url = "https://api.edenai.run/v3/chat/completions"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "openai/gpt-latest",
    "messages": [
        {"role": "user", "content": "Hello! How are you?"}
    ]
}

response = requests.post(url, headers=headers, json=payload)
result = response.json()
print(result["choices"][0]["message"]["content"])

```

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

### Python OCR request using requests library

Source: https://www.edenai.co/docs/v3/expert-models/features/ocr/ocr

Sends a POST request to the Eden AI Universal API with the OCR model (ocr/ocr/amazon) and a file UUID or URL. Requires an API key and a valid file input. Prints the JSON response.

```python
import requests

url = "https://api.edenai.run/v3/universal-ai"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "ocr/ocr/amazon",
    "input": {
        "file": "YOUR_FILE_UUID_OR_URL",
        "language": "en"
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

```

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

### Make a chat completion request with cURL

Source: https://www.edenai.co/docs/v3/quickstart/first-llm-call

Sends a chat completion request to Eden AI's OpenAI-compatible endpoint using cURL. Replace YOUR_API_KEY with your actual token. The request uses the openai/gpt-latest model.

```bash
curl -X POST https://api.edenai.run/v3/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-latest",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

```

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

### cURL: Generate image with curl

Source: https://www.edenai.co/docs/v3/expert-models/features/image/generation

Sends a POST request to the Eden AI universal endpoint using cURL with the image generation model and input text. Requires an API key.

```bash
curl -X POST https://api.edenai.run/v3/universal-ai \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "image/generation/bytedance",
    "input": {"text": "The quick brown fox jumps over the lazy dog."}
  }'

```

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

### Python: Make a Universal AI call for text moderation

Source: https://www.edenai.co/docs/v3/quickstart/first-expert-model-call

Use the Universal AI endpoint to perform text moderation with the Google provider. Replace YOUR_API_KEY with your actual token. The response is printed as JSON.

```Python
import requests

url = "https://api.edenai.run/v3/universal-ai"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "text/moderation/google",
    "input": {
        "text": "This is a sample text to moderate for harmful content."
    }
}

response = requests.post(url, headers=headers, json=payload)
result = response.json()
print(result)

```

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

### JavaScript: Make a Universal AI call for text moderation

Source: https://www.edenai.co/docs/v3/quickstart/first-expert-model-call

Use the Universal AI endpoint with fetch to perform text moderation with the Google provider. Replace YOUR_API_KEY with your actual token. The response is logged to the console.

```JavaScript
const response = await fetch('https://api.edenai.run/v3/universal-ai', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    model: 'text/moderation/google',
    input: { text: 'This is a sample text to moderate for harmful content.' }
  })
});

const result = await response.json();
console.log(result);

```

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

### cURL OCR request using curl

Source: https://www.edenai.co/docs/v3/expert-models/features/ocr/ocr

Sends a POST request to the Eden AI Universal API with the OCR model (ocr/ocr/amazon) and a file UUID or URL. Requires an API key and a valid file input. Returns the JSON response.

```bash
curl -X POST https://api.edenai.run/v3/universal-ai \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "ocr/ocr/amazon",
    "input": {"file": "YOUR_FILE_UUID_OR_URL", "language": "en"}
  }'

```

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

### POST /v3/universal-ai/async speech-to-text request

Source: https://www.edenai.co/docs/v3/expert-models/features/audio/speech-to-text-async

Sends an asynchronous speech-to-text request to the Eden AI universal API. The response contains a job ID; poll the async job endpoint until completion. Replace YOUR_API_KEY and YOUR_FILE_UUID_OR_URL with your credentials and audio file.

```python
import requests

url = "https://api.edenai.run/v3/universal-ai/async"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "audio/speech_to_text_async/amazon",
    "input": {
        "file": "YOUR_FILE_UUID_OR_URL",
        "language": "en"
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

```

```bash
curl -X POST https://api.edenai.run/v3/universal-ai/async \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "audio/speech_to_text_async/amazon",
    "input": {"file": "YOUR_FILE_UUID_OR_URL", "language": "en"}
  }'

```

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

### cURL: Text-to-Speech with curl

Source: https://www.edenai.co/docs/v3/expert-models/features/audio/tts

Sends a POST request to the Eden AI Universal AI endpoint to synthesize speech from text using the amazon/neural model. Requires an API key. The response is printed to stdout.

```bash
curl -X POST https://api.edenai.run/v3/universal-ai \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "audio/tts/amazon/neural",
    "input": {"text": "The quick brown fox jumps over the lazy dog."}
  }'

```

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

### POST /v3/universal-ai (sync) - OCR

Source: https://www.edenai.co/docs/v3/expert-models/features/ocr/ocr

Performs OCR on a PDF or image file. The file can be provided as a file ID from /v3/upload or a direct file URL. The language field is optional and may be ignored by providers that auto-detect language.

```APIDOC
## POST /v3/universal-ai (sync)

### Description
Performs OCR on a PDF or image file. The file can be provided as a file ID from /v3/upload or a direct file URL. The language field is optional and may be ignored by providers that auto-detect language.

### Method
POST

### Endpoint
/v3/universal-ai

### Parameters
#### Request Body
- **model** (string) - Required - Model string pattern: `ocr/ocr/{provider}[/{model}]`
- **input.file** (file_input) - Required - PDF or image file ID from /v3/upload or direct file URL
- **input.language** (string) - Optional - Document language code. Most providers detect the language themselves and ignore this field; providers that cannot detect it reject requests that omit it. Check the provider’s supported languages before relying on it.

### Request Example
```json
{
  "model": "ocr/ocr/amazon",
  "input": {
    "file": "YOUR_FILE_UUID_OR_URL",
    "language": "en"
  }
}
```

### Response
#### Success Response (200)
- **text** (string) - Required - The extracted text.
- **bounding_boxes** (array[object]) - Optional - List of bounding boxes.
  - **text** (string) - Required - Text within the bounding box.
  - **left** (float) - Required - Left coordinate.
  - **top** (float) - Required - Top coordinate.
  - **width** (float) - Required - Width of the box.
  - **height** (float) - Required - Height of the box.

#### Response Example
```json
{
  "text": "Extracted text",
  "bounding_boxes": [
    {
      "text": "word",
      "left": 0.1,
      "top": 0.2,
      "width": 0.3,
      "height": 0.4
    }
  ]
}
```

### Available Providers
| Provider | Model String | Price |
|---|---|---|
| amazon | `ocr/ocr/amazon` | $1.5 per 1,000 pages |
| api4ai | `ocr/ocr/api4ai` | $3 per 1,000 requests |
| google | `ocr/ocr/google` | $1.5 per 1,000 pages |
| microsoft | `ocr/ocr/microsoft` | $1 per 1,000 pages |
| mistral | `ocr/ocr/mistral` | $1 per 1,000 pages |
| sentisight | `ocr/ocr/sentisight` | $0.75 per 1,000 files |

### Quick Start
#### Python
```python
import requests

url = "https://api.edenai.run/v3/universal-ai"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "ocr/ocr/amazon",
    "input": {
        "file": "YOUR_FILE_UUID_OR_URL",
        "language": "en"
    }
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())
```

#### cURL
```bash
curl -X POST https://api.edenai.run/v3/universal-ai \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "ocr/ocr/amazon",
    "input": {"file": "YOUR_FILE_UUID_OR_URL", "language": "en"}
  }'
```
```

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

### cURL: Submit deepfake detection job

Source: https://www.edenai.co/docs/v3/expert-models/features/video/deepfake-detection-async

Sends a POST request to the async endpoint using cURL with the same model and file input. The response contains a job ID that must be polled until completion.

```bash
curl -X POST https://api.edenai.run/v3/universal-ai/async \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "video/deepfake_detection_async/resemble",
    "input": {"file": "YOUR_FILE_UUID_OR_URL"}
  }'

```

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

### cURL: Make a Universal AI call for text moderation

Source: https://www.edenai.co/docs/v3/quickstart/first-expert-model-call

Use the Universal AI endpoint with cURL to perform text moderation with the Google provider. Replace YOUR_API_KEY with your actual token.

```cURL
curl -X POST https://api.edenai.run/v3/universal-ai \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "text/moderation/google",
    "input": {
      "text": "This is a sample text to moderate for harmful content."
    }
  }'

```
Tokens:
Raw