Bland BTTS is available on Novita AI through a native audio API that forwards requests to Bland’s production speech engine. It fits teams building phone agents that want Bland’s voices, free voice cloning, and telephony-ready audio while authenticating with a single Novita AI API key. Synthesis costs $0.04 per 1,000 characters, voice cloning is free, and both endpoints return synchronously.
Key Takeaways
- Bland BTTS is live on Novita AI as a passthrough proxy to Bland’s engine, billed through your Novita account.
- Voice cloning is free. Only synthesis is billed, at $0.04 per 1,000 characters; Han characters count as two.
- It is built for speech agents, with telephony-oriented audio output and expression controls on BTTS V3 voices.
- Requests are synchronous and capped at 5,000 characters, which suits short utterances rather than long scripts; voices are referenced by UUID rather than display name.
Bland BTTS Specs and Pricing Summary
| Field | Details |
|---|---|
| Display name | Bland BTTS |
| Model slug | bland-tts |
| Base URL | https://api.novita.ai |
| Endpoints | POST /v3/bland-tts/v2/tts for speech synthesis; POST /v3/bland-tts/v1/voices/clone for voice cloning |
| Request limit | 5,000 characters of text per synthesis request |
| Output | Audio binary, not JSON |
| Voice controls | Expressiveness and stability controls on BTTS V3 voices |
| Pricing | $0.04 per 1,000 characters (SKU BLAND_BTTS); voice cloning is free |
| Best fit | Phone agents and voice products that need Bland’s engine and telephony audio |
Specs from the Novita AI model page and API reference; pricing from the Novita AI pricing page, checked October 23, 2026.
What Is Bland BTTS?
Bland BTTS is the text-to-speech engine from Bland AI, the company known for voice agents that handle phone calls. It turns text into audio and clones voices from short samples. BTTS has two selectable generations, V2 and V3, and the expressiveness controls apply only to V3 voices.
The design reflects Bland’s phone-agent focus: audio returns as binary data rather than a JSON job, the output path supports telephony use, and a clone needs only a short sample.
On Novita AI, Bland BTTS is a passthrough proxy, not a natively hosted model. Novita forwards requests to Bland’s engine and bills them through your Novita account, so the voices and the native request schema are Bland’s. That differs from hosted audio models such as Fish Audio S2 Pro, which Novita serves directly. If you specifically want Bland’s voice stack, the proxy saves a second vendor integration.
Bland BTTS API Access on Novita AI
The Bland TTS API on Novita AI exposes two endpoints. Speech synthesis accepts text and a voice UUID and returns audio bytes. Voice cloning accepts a short audio sample and returns a JSON voice_id. Both use your Novita AI API key, and neither takes a model field: routing is determined by the path.
Bland BTTS uses its own request schema rather than an OpenAI-compatible one, so an existing Novita text integration needs a separate client path. The Bland BTTS API reference documents the full request and response formats.
Voices are referenced by UUID, not by display name. Cloned voices are shared among all callers using the same API key, and the API does not record which caller created which voice, so teams sharing a key need their own ownership tracking.
Key Capabilities for Developers
Free voice cloning changes onboarding economics
Bland voice cloning is free on Novita AI. The clone call takes a short audio sample and returns a UUID that synthesis can reference, and the call itself generates no charge. That differs from voice APIs that charge a one-time fee per clone before any synthesis is billed, and it means a catalog of branded voices costs only the seconds those voices speak.
Built for telephony, with V3-only expression controls
Bland’s engine is designed for phone conversations, and BTTS V3 voices add expressiveness and stability controls that V2 voices do not have. If you need multilingual cloning or emotion tags for studio-style work, a hosted model is the closer fit.
Synchronous responses and a 5,000-character ceiling
A successful request returns audio directly, so there is no task ID to poll. The tradeoff is a 5,000-character limit per request, which covers most agent turns but makes the API a poor fit for long-form narration.
When to Use Bland BTTS
- Phone and voice agents that need Bland’s voices or telephony-oriented output
- Products already on Novita AI that want Bland’s engine without a separate Bland integration
- Workflows with many branded voices, where free cloning keeps onboarding cost low
- Turn-by-turn synthesis inside a conversation, where a synchronous response is simpler than a job queue
When Not to Use Bland BTTS
- If you need one OpenAI-compatible SDK for both text and speech, the native protocol adds a separate client path
- If you need long-form narration in a single request, the 5,000-character cap makes it a poor fit
- If multiple tenants share one API key, the shared voice library does not isolate them
- If you want a hosted model with multilingual cloning and emotion tags, Fish Audio S2 Pro is the closer fit
How Bland BTTS Fits Your API Workflow
Bland BTTS fits as the speech layer behind a phone agent or voice product: the agent sends text and receives audio it can play into the call path. Synthesis is synchronous and capped at 5,000 characters, which matches turn-by-turn prompts but not long scripts. Billing follows the same shape, since only synthesis is charged per character while voice creation stays free.
Conclusion
Bland BTTS on Novita AI is the shortest route to Bland’s speech engine if you are already on Novita’s API surface. You get Bland’s V2 and V3 voices, free voice cloning, telephony-ready audio, and expression controls on V3 voices through two endpoints and a Novita API key. The constraints are equally clear: a 5,000-character request cap, UUID-only voice references, and a voice library shared per key.
If your workload is phone agents, IVR prompts, or branded voice libraries, Bland BTTS is the direct fit. If you need long-form narration or an OpenAI-style speech SDK, compare hosted alternatives first, including the options in our text-to-speech API comparison.
FAQ
Is Bland AI TTS available on Novita AI?
Yes. The model is listed as Bland BTTS on Novita AI, with speech synthesis and voice cloning endpoints. The model page opens the playground and links to the current API guide.
How much does Bland BTTS cost on Novita AI?
On Novita AI, Bland BTTS synthesis costs $0.04 per 1,000 characters, and Han characters count as two. Voice cloning is free. Check the Novita AI pricing page before budgeting production volume.
Does voice cloning cost anything on Novita AI?
No. Bland BTTS voice cloning generates no charge. The clone call takes a short audio sample and returns a voice UUID that synthesis can use. The Bland BTTS API reference covers voice creation.
Does Bland BTTS return audio directly or a task ID on Novita AI?
Bland BTTS synthesis returns audio binary directly, not a JSON task to poll. The model page links to the API reference for the current output options.
Can I use Bland BTTS for phone agents on Novita AI?
Yes, that is its strongest fit. Bland BTTS is designed for speech agents, with telephony-oriented audio output and expression controls on BTTS V3 voices. The Bland BTTS API reference covers the endpoints.
