ElevenLabs
Operational· checked every few minutes

Eleven Flash v2.5 API

The Eleven Flash v2.5 API: ElevenLabs' ultra-fast speech model for real-time use, in 32 languages. $0.05 per 1K characters, on the same key as every other model.

Price of speech, by the text you send
1K characters
$0.05
A minute of speech
$0.043
Model IDeleven-flash-v2.5Updated
speak.py
from openai import OpenAI
client = OpenAI(
base_url="https://api.zurelay.com/v1",
api_key="YOUR_ZURELAY_KEY",
)
speech = client.audio.speech.create(
model="eleven-flash-v2.5",
voice="Talia",
input="Welcome to Zurelay. Your video is ready.",
)
speech.write_to_file("welcome.mp3")

Base URL https://api.zurelay.com/v1. Works with any OpenAI SDK. Speech API docs

Voices
17, and OpenAI's names
Speed
0.7x to 1.2x
Formats
MP3 · WAV · PCM
Text per request
Up to 5,000 characters
A minute of speech
About $0.043
Released
Dec 18, 2024
Status
Operational

Pricing

Eleven Flash v2.5 API pricing. By the character.

Pay as you go from prepaid credit, with no subscription. Top up from $10. Requests that fail are never billed.

RateZurelayElevenLabs list
1K characters
of the text you send
$0.05$0.04
A minute of speech
about 850 characters
$0.043$0.034

Billed on the characters of text you send, spaces and punctuation included; the audio is included in the price. A minute of speech is about 850 characters. Requests that fail cost nothing.

Every ElevenLabs model and price
Cost calculator

Move the slider to the text you speak in a month.

Characters per month250K
A month on Zurelay
$12.50
Speech you get
about 4.9 hours

Billed on the characters you send, spaces and punctuation included. Requests that fail cost nothing.

Cost

Eleven Flash v2.5 API cost, worked out.

Typical workloads at Zurelay's live prices.

  • A voice message

    About a minute of speech.

    $0.043
  • A podcast episode

    About 30 minutes of speech.

    $1.29
  • An audiobook

    About 10 hours of speech.

    $25.80

Overview

What is Eleven Flash v2.5?

Eleven Flash v2.5 is ElevenLabs' ultra-fast speech model for real-time use, released in December 2024, in 32 languages. The Zurelay Eleven Flash v2.5 API serves it through an OpenAI-compatible speech endpoint with 17 ElevenLabs voices, at $0.05 per 1,000 characters of text, about $0.043 a minute of speech.

ElevenLabs announced Eleven Flash v2.5 on December 18, 2024 as an ultra-fast model optimized for real-time use, and says it answers in about 100 milliseconds on its own servers when streaming. It speaks 32 languages. On Zurelay its ID is eleven-flash-v2.5, and ElevenLabs' own ID, eleven_flash_v2_5, works too.

Eleven Flash v2.5 takes no style instructions and reads the text as written. Set the pace with speed, from 0.7 to 1.2. Measured on Zurelay, the same sentence took 6.7 seconds at 0.7 and 4.0 seconds at 1.2. ElevenLabs leaves text normalization off on this model, so write numbers and abbreviations the way they should be said.

Zurelay serves it through an OpenAI-compatible endpoint: POST https://api.zurelay.com/v1/audio/speech with up to 5,000 characters of input and one of 17 ElevenLabs voices (Talia by default, or an OpenAI name such as alloy or nova). The reply is the audio file itself: MP3 by default, or WAV or raw 24 kHz PCM. A sentence usually comes back whole in under a second. Each request speaks in one voice. Voice cloning, ElevenLabs' voice library, multi-speaker dialogue and streaming aren't available through Zurelay.

Speech on this model is billed on the text alone: $0.05 per 1,000 characters you send, spaces, punctuation and any tags included, with the audio included whatever its length. A minute of speech is about 850 characters, so about $0.043. ElevenLabs' own API price is $0.04 per 1,000 characters. Each reply's x-zurelay-input-tokens header shows the characters it was billed on, every request is in your request log, and requests that fail cost nothing.

Strengths

Where Eleven Flash v2.5 shines. And what teams build with it.

01

The quickest replies

Built for real-time use: on Zurelay a sentence comes back whole in under a second.

02

Half the price

Costs half of what Eleven v4 and Multilingual v2 do for the same text.

03

A pace you set

speed from 0.7 to 1.2 slows the reading down or picks it up.

Use cases

  • Voice assistants

    Replies spoken a sentence at a time, with little wait.

  • Notifications and alerts

    Order updates, reminders and alerts read out as they happen.

  • Reading apps

    Articles and messages read aloud on demand.

  • High-volume speech

    Lots of short clips at ElevenLabs' lowest price per character.

Get started

Call Eleven Flash v2.5 in three steps.

No waitlist and no new SDK. If your code already talks to OpenAI, it already talks to Zurelay.

  1. 1

    Create a key

    Sign up, add credit and create an API key. Set a monthly budget or a rate limit per key if you like.

  2. 2

    Point your SDK at Zurelay

    Change the base URL. Everything else in your code stays the same.

    https://api.zurelay.com/v1
  3. 3

    Use eleven-flash-v2.5

    POST your text and a voice to /v1/audio/speech and get the audio back as MP3, WAV or PCM.

    eleven-flash-v2.5

FAQ

Eleven Flash v2.5 API questions.

Anything else? Ask our team and we’ll answer by email.

How much does the Eleven Flash v2.5 API cost?

Zurelay charges $0.05 per 1,000 characters of the text you send, spaces and punctuation included, and the audio is included in that. A minute of speech is about 850 characters, so about $0.043. ElevenLabs' own API price for Eleven Flash v2.5 is $0.04 per 1,000 characters: on Zurelay you pay a little more and get it on the same key and prepaid balance as every other model, with no ElevenLabs subscription. Requests that fail cost nothing.

How do I call Eleven Flash v2.5 with the OpenAI SDK?

Create a Zurelay API key, set the SDK's base URL to https://api.zurelay.com/v1 and call client.audio.speech.create with model set to eleven-flash-v2.5, a voice such as Talia and your text as input. Leave instructions out: this model doesn't take them. The reply is the audio file itself: save it with write_to_file in Python, or write its bytes to a file in Node. To try voices first, use the Speech mode in the Playground at https://zurelay.com/app/playground.

Which voices does Eleven Flash v2.5 have on Zurelay?

17 ElevenLabs voices: Talia, Elara, Baxter, Eldrin, Elowen, Kaelen, Lawrence, Alicia, Maisie, Warren, Jade, Eddie, Caleb, Sawyer, Finley, Florence and Darian, with Talia as the default. OpenAI voice names work too and map to one of them: alloy speaks as Elowen, nova as Elara, onyx as Sawyer and shimmer as Florence. Gemini's voice names, such as Kore, are refused on this model. The earlier names, such as Sarah or Brian, still work and speak with the voice ElevenLabs replaced them with. Voice cloning, voice design and ElevenLabs' voice library aren't available through Zurelay.

Can I change how fast Eleven Flash v2.5 speaks?

Yes. Set speed between 0.7 and 1.2, where 1 is the normal pace. It takes no style instructions, though: a request with instructions is refused with an unsupported_parameter error, so the mood comes from your words and punctuation.

Which languages does Eleven Flash v2.5 speak?

32: the 29 languages of Eleven Multilingual v2, plus Hungarian, Norwegian and Vietnamese. There's no language setting on Zurelay: write the text in the language you want spoken.

What audio formats and text limits does Eleven Flash v2.5 have?

Each request takes up to 5,000 characters of input on Zurelay; split longer text into several requests. Audio comes back as MP3 by default, or set response_format to wav, or to pcm for raw 24 kHz, 16-bit mono samples. Each request speaks in one voice, and the whole file comes back in one reply. Zurelay doesn't store the audio, so save the file you get back.

Is this ElevenLabs' Eleven Flash v2.5?

Yes. Your requests run on ElevenLabs' Eleven Flash v2.5 (eleven_flash_v2_5) through ElevenLabs' own API, and Zurelay doesn't swap in a different model. Zurelay is an independent service and is not affiliated with ElevenLabs.

More models on Zurelay

All models

Stop paying list price.
Start saving today.

Create a key in seconds, point the SDK you already use at Zurelay, and every request costs up to 90% less from the first token.