# ElevenLabs is now on OpenRouter

By Steven Van · 2026-10-07

All OpenRouter customers get 50% off the 11 ElevenLabs models through October 19 at 8am PT, with no separate ElevenLabs plan required.

Apps that already use [OpenRouter](<https://creatorstoolbox.com/tools/openrouter>) can now add speech generation and transcription using the same API key, without a separate ElevenLabs plan. The [launch](<https://openrouter.ai/blog/announcements/elevenlabs-on-openrouter/>) brings nine Text to Speech models and two Speech to Text models.

## Choose a model for the job

Eleven v4 supports expressive narration and character work, with audio tags such as \[whispering\] and \[laughing\] to direct delivery. Eleven v4 Turbo targets low-latency voice agent replies and bills at half the v4 per-character rate.

Multilingual v2 offers speed control for long-form audio. Flash v2.5 supports 32 languages and up to 40,000 characters per request for high-volume speech. The lineup also includes v3, v3 Conversational, Flash v2, Turbo v2.5 and Turbo v2.

For transcription, Scribe v2 provides word timestamps, speaker labels and audio-event tags in more than 90 languages. Scribe v2 Medical is fine-tuned for medical terminology, with ElevenLabs reporting 35% fewer errors on clinical audio than Scribe v2. Both use the same request format and per-second price. Realtime Scribe v2 over WebSocket is not included in this launch.

## From narration to spoken replies

Speech generation uses POST /api/v1/audio/speech, while transcription uses POST /api/v1/audio/transcriptions. Speech requests accept any of ElevenLabs’ 21 premade voices by name and an explicit output format, such as MP3.

The walkthrough starts with a product tour narrated by v4. Longer scripts are split at paragraph boundaries to fit its 10,000-character request limit. Passing neighboring text through ElevenLabs-specific options helps keep intonation consistent between chunks; audio tags count toward billed characters.

A voice agent then chains Scribe v2 to transcribe a spoken question, a chat model to answer it, and v4 Turbo to generate the reply. The chat model can be changed without altering the audio calls. Scribe v2 also supports speaker-labeled transcripts for the team’s review meeting.

## Launch pricing

Every ElevenLabs model is **50% off OpenRouter’s list price** for all OpenRouter customers through October 19 at 8am PT.

![OpenRouter](<https://openrouter.ai/favicon/glyph.png>)

OpenRouter

One API for 500+ AI models across 80+ providers — pay with credits that work anywhere, with automatic fallback when a provider goes down.

[View OpenRouter →](<https://creatorstoolbox.com/tools/openrouter>)

[Original source](<https://openrouter.ai/blog/announcements/elevenlabs-on-openrouter/>)

[Read ElevenLabs is now on OpenRouter on Creators Toolbox](<https://creatorstoolbox.com/blog/openrouter-elevenlabs-is-now-on-openrouter>)

---
Canonical source: https://creatorstoolbox.com/blog/openrouter-elevenlabs-is-now-on-openrouter
