Technical Specifications of tts-1
| Specification | Details |
|---|---|
| Model ID | tts-1 |
| Provider | OpenAI |
| Model type | Text-to-speech (TTS) model for converting text input into spoken audio. |
| Primary optimization | Optimized for speed and low-latency generation, especially for realtime or near-realtime speech output. |
| Quality profile | Lower latency than tts-1-hd, but with lower audio quality than the HD variant. |
| Input modality | Text only. |
| Output modality | Audio only. |
| API endpoint | OpenAI Audio API speech generation endpoint: /v1/audio/speech. |
| Max input length | Up to 4096 characters per request. |
| Supported response formats | mp3, opus, aac, flac, wav, pcm. |
| Speed control | Supported from 0.25 to 4.0, with 1.0 as default. |
Voice options for tts-1 | alloy, ash, coral, echo, fable, onyx, nova, sage, shimmer. |
| Streaming support | The Speech API supports streaming audio output, but SSE streaming and instruction-based voice control are not supported for tts-1. |
| Pricing | OpenAI lists tts-1 at $15 per 1M tokens for speech generation. |
What is tts-1?
tts-1 is OpenAIโs text-to-speech model designed to turn written text into natural-sounding spoken audio. It is positioned as the faster, lower-latency option among OpenAIโs classic TTS models, making it suitable for applications that need quick speech synthesis rather than the highest possible fidelity.
Developers typically use tts-1 through the Audio APIโs speech generation endpoint when they want to convert application text, prompts, notifications, narrations, or assistant responses into playable audio files. OpenAIโs documentation describes it as optimized for realtime text-to-speech use cases.
In practice, tts-1 is a good fit for lightweight voice experiences, rapid response systems, interactive prototypes, and products where responsiveness matters more than premium voice quality. If maximum quality is the priority, OpenAI points users toward tts-1-hd, while newer expressive use cases may use newer TTS models instead.
Main features of tts-1
- Low-latency speech generation:
tts-1is specifically optimized for speed, which makes it useful for apps that need spoken output quickly. - Natural-sounding text-to-speech: The model converts plain text into spoken audio suitable for narration, assistant responses, and voice interfaces.
- Multiple built-in voices:
tts-1supports a set of built-in voices including alloy, ash, coral, echo, fable, onyx, nova, sage, and shimmer. - Flexible audio output formats: Developers can request generated audio in common formats such as MP3, WAV, FLAC, AAC, Opus, and PCM depending on playback or processing needs.
- Adjustable playback speed: The API allows speed control from 0.25x to 4.0x, enabling slower narration or faster playback where appropriate.
- Simple API-based integration:
tts-1is available through the standard speech generation API, which makes it straightforward to integrate into web, mobile, or backend workflows. - Good for realtime-oriented applications: OpenAI explicitly frames
tts-1as a model for realtime text-to-speech scenarios, which makes it practical for assistants, notifications, and fast interactive systems. - Tradeoff-focused model choice: Compared with
tts-1-hd, this model prioritizes faster generation over higher-fidelity output, giving developers a clear latency-versus-quality option.
How to access and integrate tts-1
Step 1: Sign Up for API Key
To access the tts-1 API, first sign up on CometAPI and generate your API key from the dashboard. After logging in, create a new key, copy it securely, and store it in your application environment variables. You will use this key to authenticate all requests to the tts-1 API.
Step 2: Send Requests to tts-1 API
Once you have your API key, send a POST request to the CometAPI endpoint for tts-1 with your input payload. Include your API key in the Authorization header and specify tts-1 as the model. A typical request includes the input text plus TTS parameters such as voice and response format.
curl https://api.cometapi.com/v1/audio/speech \
-H "Authorization: Bearer $COMETAPI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "tts-1",
"input": "Welcome to CometAPI text to speech.",
"voice": "alloy",
"response_format": "mp3"
}' \
--output speech.mp3
Step 3: Retrieve and Verify Results
After submitting your request, the tts-1 API returns generated audio content if the call succeeds. Save the returned file or stream, verify that the audio plays correctly, and confirm that the selected voice, speed, and format match your application requirements. If needed, retry with adjusted parameters to improve the final output.