Name: Sarvam AI – Indic Translation & Speech | Free Plan | API.market
Brand: sarvam
SKU: api-sarvam-ai-models
Availability: InStock
Rating: 3.00 (1 reviews)
Author: API Market

Shashank Agarwal

### Sarvam AI Model API Suite

The **Sarvam AI Model API Suite** is a comprehensive set of APIs designed to bring advanced natural language processing and speech capabilities to your applications. Ideal for developers and enterprises, the suite includes APIs for text translation, speech-to-text transcription, speech-to-text translation, call analytics, text analytics, and text-to-speech conversion. This suite enables seamless integration of sophisticated AI-driven functionalities, enhancing user interactions, accessibility, and analytics for applications in customer support, content creation, media, and real-time language processing.

#### Key Features
- **Text Translation**: Supports translating text across various Indic languages, with options for formal, code-mixed, and gender-specific translations.
- **Speech-to-Text**: Converts spoken audio input into text, enabling transcription for a range of Indic languages with optional word-level timestamps.
- **Speech-to-Text Translation**: Directly transcribes and translates audio to English in a single operation, with automatic language detection.
- **Call Analytics**: Analyzes call audio to provide structured responses to specific questions, with reasoning and direct quotes from the conversation.
- **Text Analytics**: Performs in-depth analysis on text content, generating answers to specified questions and providing reasoning and source references.
- **Text-to-Speech**: Converts text to natural-sounding spoken audio in various voices, with adjustable pitch, pace, loudness, and speaker options.

#### Supported Languages
- **Text Translation, Speech-to-Text, and Text-to-Speech**: Supported languages include Hindi (`hi-IN`), Bengali (`bn-IN`), Kannada (`kn-IN`), Malayalam (`ml-IN`), Marathi (`mr-IN`), Odia (`od-IN`), Punjabi (`pa-IN`), Tamil (`ta-IN`), Telugu (`te-IN`), Gujarati (`gu-IN`), and English (`en-IN`).

---

### Use Cases
1. **Customer Support Automation**: Use Speech-to-Text, Call Analytics, and Text Analytics APIs to streamline support, analyze interactions, and gather insights on customer issues.
2. **Educational Language Translation**: Empower educators and students with Text Translation and Text-to-Speech to create multilingual content and promote accessibility.
3. **Media Content Creation**: Utilize Text-to-Speech to create audio content for videos, podcasts, and multilingual channels.
4. **Business Analytics for Call Centers**: Gain deep insights into customer interactions with Call Analytics and Text Analytics, optimizing service and identifying trends.
5. **Voice-Enabled LLM Applications**: Build voice-based large language model applications using Speech-to-Text Translation and Text-to-Speech for real-time interaction and translation.

---

### API Endpoints & Detailed Examples

---

#### 1. **Text Translation API**
**Endpoint**: `POST https://api.magicapi.dev/api/v1/sarvam/ai-models/translate`

**Description**: Translates input text to the target language, supporting formal and code-mixed styles with gender-specific customization.

**Example Request**:
```bash
curl -X 'POST' \
  'https://api.magicapi.dev/api/v1/sarvam/ai-models/translate' \
  -H 'accept: application/json' \
  -H 'x-magicapi-key: API_MARKET_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "input": "This is a sample string to translate to Hindi",
  "source_language_code": "en-IN",
  "target_language_code": "hi-IN",
  "speaker_gender": "Male",
  "mode": "formal",
  "model": "mayura:v1",
  "enable_preprocessing": false
}'
```

**Example Response**:
```json
{
  "translated_text": "यह एक नमूना वाक्य है जो हिंदी में अनुवाद करने के लिए है।"
}
```

---

#### 2. **Speech-to-Text API**
**Endpoint**: `POST https://api.magicapi.dev/api/v1/sarvam/ai-models/speech-to-text`

**Description**: Converts audio input to text, supporting a variety of Indic languages with optional word-level timestamps.

**Example Request**:
```bash
curl -X 'POST' \
  'https://api.magicapi.dev/api/v1/sarvam/ai-models/speech-to-text' \
  -H 'accept: application/json' \
  -H 'x-magicapi-key: API_MARKET_API_KEY' \
  -H 'Content-Type: multipart/form-data' \
  -F 'file=@audio_sample.wav' \
  -F 'language_code=hi-IN' \
  -F 'model=saarika:v1' \
  -F 'with_timestamps=true'
```

**Example Response**:
```json
{
  "transcript": "Agent: धन्यवाद! ग्राहक सहायता में आपका स्वागत है।",
  "timestamps": {
    "words": ["धन्यवाद", "ग्राहक", "सहायता", "में", "आपका", "स्वागत"],
    "start_time_seconds": [0.5, 1.2, 2.3, 3.1, 4.0, 5.0],
    "end_time_seconds": [1.1, 2.2, 3.0, 3.8, 4.9, 5.8]
  }
}
```

---

#### 3. **Speech-to-Text Translation API**
**Endpoint**: `POST https://api.magicapi.dev/api/v1/sarvam/ai-models/speech-to-text-translate`

**Description**: Automatically detects the language in an audio file, transcribes it, and translates the output to English.

**Example Request**:
```bash
curl -X 'POST' \
  'https://api.magicapi.dev/api/v1/sarvam/ai-models/speech-to-text-translate' \
  -H 'accept: application/json' \
  -H 'x-magicapi-key: API_MARKET_API_KEY' \
  -H 'Content-Type: multipart/form-data' \
  -F 'file=@audio_sample.wav' \
  -F 'prompt=Create a nice translation like a normal person' \
  -F 'model=saaras:v1'
```

**Example Response**:
```json
{
  "transcript": "Hello, welcome to customer support.",
  "language_code": "hi-IN"
}
```

---

#### 4. **Call Analytics API**
**Endpoint**: `POST https://api.magicapi.dev/api/v1/sarvam/ai-models/call-analytics`

**Description**: Analyzes call audio to answer predefined questions with explanations and direct quotes from the call.

**Example Request**:
```bash
curl -X 'POST' \
  'https://api.magicapi.dev/api/v1/sarvam/ai-models/call-analytics' \
  -H 'accept: application/json' \
  -H 'x-magicapi-key: API_MARKET_API_KEY' \
  -H 'Content-Type: multipart/form-data' \
  -F 'file=@call_audio.mp3' \
  -F 'questions=What was the main point discussed?' \
  -F 'hotwords=sales, product, AI, API'
```

**Example Response**:
```json
{
  "answers": [
    {
      "id": "q1",
      "question": "What was the main point discussed?",
      "reasoning": "The customer mentioned concerns about API integration.",
      "response": "API integration issues",
      "utterance": "I'm having trouble integrating the API with our system."
    }
  ],
  "duration_in_seconds": 180.5,
  "file_name": "call_audio.mp3",
  "transcript": "Customer: I'm having trouble integrating the API with our system."
}
```

---

#### 5. **Text Analytics API**
**Endpoint**: `POST https://api.magicapi.dev/api/v1/sarvam/ai-models/text-analytics`

**Description**: Analyzes provided text content and answers specific questions about the text with supporting reasoning.

**Example Request**:
```bash
curl -X 'POST' \
  'https://api.magicapi.dev/api/v1/sarvam/ai-models/text-analytics' \
  -H 'accept: application/json' \
  -H 'x-magicapi-key: API_MARKET_API_KEY' \
  -H 'Content-Type: application/x-www-form-urlencoded' \
  --data-urlencode 'text=Climate change is a pressing global issue that demands attention.' \
  --data-urlencode 'questions=What is the main topic of this text?'
```

**Example Response**:
```json
{
  "answers": [
    {
      "id": "q1",
      "question": "What is the main topic of this text?",
      "reasoning": "The text focuses on climate change.",
      "response": "Climate change",
      "utterance": "Climate change is a pressing global issue."
    }
  ]
}
```

---

#### 6. **Text-to-Speech API**
**Endpoint**: `POST https://api.magicapi.dev/api/v1/sarvam/ai-models/text-to-speech`

**Description**: Converts text into spoken audio, supporting multiple Indic voices with customizable parameters like pitch, pace, and loudness.

**Example Request**:
```bash
curl -X 'POST' \
  'https://api.magicapi.dev/api/v1/sarvam/ai-models/text-to-speech' \
  -H 'accept: application/json' \
  -H 'x-magicapi-key: API_MARKET_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "

inputs": ["The yellow fox jumped over the fence.", "Another example text."],
  "target_language_code": "hi-IN",
  "speaker": "meera",
  "pitch": 0,
  "pace": 1,
  "loudness": 1,
  "speech_sample_rate": 16000,
  "enable_preprocessing": true,
  "model": "bulbul:v1"
}'
```

**Example Response**:
```json
{
  "audios": ["<base64-encoded audio string>"]
}
```

---

### Authentication

Each API request must include the `x-magicapi-key` in the header. Register and obtain an API key from [Sarvam AI](https://api.magicapi.dev/api/v1/sarvam/ai-models) to start using the suite.

---

### Why Choose Sarvam AI Model API Suite?

The Sarvam AI Model API Suite provides a robust and flexible solution for language and speech processing, with customizable parameters, support for multiple Indic languages, and a wide range of use cases from customer support to media and analytics. Integrate Sarvam AI into your applications to leverage advanced AI capabilities and create rich, interactive experiences.

Translate, transcribe, and synthesize voice in Indian languages. Supports code-mixed input. Free 1000 calls/mo. Start free on API.market.

Sarvam AI – Indic Translation & Speech | Free Plan | API.market

FREE

Sarvam

Works well, but missing a higher Paid plan.