# Eleven Flash v2.5

> ElevenLabs' low-latency Flash v2.5 text-to-speech model, available as a partner-hosted SKU on Cloudflare Workers AI. Tuned for fast multilingual speech generation from text.

**ID**: `elevenlabs/eleven-flash-v2-5`  
**Creator**: elevenlabs  
**Category**: audio  
**Context**: Not published  
**Status**: Disabled (since 2026-07-26) — no longer routable or listed; kept for history/usage reference.  
**Reason**: The catalogue was narrowed to models whose output is text or embeddings; this model's output is audio (text-to-speech), so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend).  
**Pricing**: $0.05 per 1K characters of input text  
**Released**: 2026-07-24  
**Web page**: https://anyrouter.dev/model/elevenlabs/eleven-flash-v2-5

**Input modalities**: text  
**Output modalities**: audio  

**Capabilities**: text-to-speech

**Aliases**: `eleven-flash-v2-5`

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/audio/speech`

```bash
curl https://anyrouter.dev/api/v1/audio/speech \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  --output speech.mp3 \
  -d '{
  "model": "elevenlabs/eleven-flash-v2-5",
  "input": "Hello! This is a text-to-speech test.",
  "voice_id": "<voice id>",
  "response_format": "mp3"
}'
```

`voice_id` is required and has no default. ElevenLabs voice ids are opaque per-account strings (e.g. "21m00Tcm4TlvDq8ikWAM"). Copy one from the ElevenLabs Voice Library — there is no default, and a request without one is rejected.

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| Cloudflare AI Workers | $0.05 per 1K characters of input text | — |

## Source

- Upstream docs: https://developers.cloudflare.com/ai/models/elevenlabs/eleven-flash-v2-5/