OpenAI Text-to-Speech
OpenAI Text-to-Speech is a cloud API that converts written text into spoken audio with the GPT-4o Mini TTS model.
What OpenAI Text-to-Speech is
OpenAI Text-to-Speech is a cloud API that converts written text into spoken audio with the GPT-4o Mini TTS model. It supports 57 languages, API access, commercial use, and voice cloning. The speech guide describes blog narration, multilingual audio, and realtime output through streaming. Instructions can influence accent, emotional range, intonation, impressions, speed, tone, and whispering. Thirteen built-in voices are listed, with marin and cedar recommended for quality; voices are currently optimized for English. Chunked streaming can begin playback before a complete file is generated. MP3 is the default output, with Opus, AAC, FLAC, WAV, and PCM also supported. Listed plans include TTS-1 at $15 a month, TTS-1 HD at $30 a month, and GPT-4o Mini TTS API usage at $0.60 a month per 1 million text input tokens, plus $12.00 per 1 million audio tokens. The latter has no free tier. The guide lists JavaScript, Python, Go, Java, Ruby, and command-line examples.
Good
- Supports 57 languages
- Offers MP3, Opus, AAC, FLAC, WAV, and PCM
- Provides streaming audio playback
- Includes voice cloning and custom voice options
- Supports multiple SDK languages
Watch out
- No free plan or free tier listed
- Voices are currently optimized for English
- Input limits are documented as 4,096 and 2,000 tokens
Who it suits
It suits developers building speech generation into cloud applications through an API. It is also relevant for projects needing multilingual output, streaming, voice controls, or commercial use.
Verdict
OpenAI Text-to-Speech provides configurable speech generation, streaming, and several audio formats through an API. Review the listed input limits, language voice optimization, and usage charges for your workflow.
Plans and prices
From OpenAI Text-to-Speech's own pricing page, checked 20 Sep 2026.
| Plan | Price | What it gives you |
|---|---|---|
| TTS-1 | $15 a month | Text-to-speech model optimized for speed · Speech generation through the Audio API |
| TTS-1 HD | $30 a month | Higher-quality text-to-speech model |
| GPT-4o Mini TTS API usage | $0.60 a month | maximum 2,000 input tokens · Free tier not supported · per 1M text input tokens; audio output billed separately at $12.00 per 1M audio tokens |
The facts
Each one read from OpenAI Text-to-Speech's own pages, with the date it was checked.
What it does
- API access
- Yeschecked 20 Sep 2026
- API integration
- The guide provides examples using OpenAI SDKs for JavaScript, Python, Go, Java, Ruby, and the command line.checked 7 Oct 2026
- Commercial use
- Yeschecked 20 Sep 2026
- Custom voices
- An approved custom voice can be created from a speaker’s consent recording and matching audio sample, subject to eligibility and recording requirements.checked 7 Oct 2026
- Data use and retention
- OpenAI says API Platform data is not used for model training by default and that API inputs and outputs may be retained for up to 30 days, with zero data retention available for eligible endpoints and qualifying use cases.checked 7 Oct 2026
- Export formats
- mp3,opus,aac,flac,wav,pcmchecked 20 Sep 2026
- Languages
- The TTS model can generate spoken audio in the languages listed in the guide by supplying text in the desired language.checked 7 Oct 2026
- Maximum input
- 4096checked 20 Sep 2026
- Model input limit
- The model page states a maximum input size of 2,000 tokens.checked 7 Oct 2026
- Output formats
- The default response format is MP3, and supported formats include Opus, AAC, FLAC, WAV, and PCM.checked 7 Oct 2026
- Security
- OpenAI says API Platform data is encrypted at rest with AES-256 and in transit with TLS 1.2 or higher, and that the API Platform is SOC 2 Type 2 audited.checked 7 Oct 2026
- Speech controls
- With gpt-4o-mini-tts, instructions can control accent, emotional range, intonation, impressions, speech speed, tone, and whispering.checked 7 Oct 2026
- Speech uses
- The guide lists narrating written blog posts, producing spoken audio in multiple languages, and realtime audio output using streaming as example uses.checked 7 Oct 2026
- Streaming
- Speech API chunked audio streaming lets playback begin before the complete file is generated.checked 7 Oct 2026
- User disclosure
- OpenAI usage policies require clear disclosure to end users that the TTS voice they hear is AI-generated and not a human voice.checked 7 Oct 2026
- Voice cloning
- Yeschecked 20 Sep 2026
- Voices
- The guide lists 13 built-in voices and recommends marin or cedar for best quality; voices are currently optimized for English.checked 7 Oct 2026
- What it does
- The GPT-4o Mini TTS model converts text into natural sounding spoken text through the speech generation API endpoint.checked 7 Oct 2026
The company
- Maker
- OpenAI describes itself as an AI research and deployment company.checked 7 Oct 2026
Where it runs
- Platforms
- apichecked 20 Sep 2026
Alternatives to OpenAI Text-to-Speech
- 1
Google Cloud Vision API
Free planWebAPILinuxmacOS7.9$0.60 usage-based - 2
Sightengine
Free planWebAPISelf-hosted7.5From $29 a month - 3
WebPurify
Free planAndroidiOSWebAPI7.4From $5 a month - 4
Amazon Rekognition Content Moderation
Free planWebAPI7.3From $0.10 a month - 5
Tisane API
Free planWebAPILinuxmacOS7.1From $99 a month - 6
Alibaba Cloud Content Moderation
iOSWebAPI7.0$0.09 usage-based