ElevenLabs Multilingual v2

Overview

This model converts text input into audio output with support for expressive voice profiles and brand-safe delivery. It operates in a balanced cost tier with medium generation speed and is designed specifically for speech synthesis. On Kyma, it runs through an OpenAI-compatible endpoint using a single API key. Prompt caching bills repeated prompt prefixes at this model’s cached input rate. Requests automatically failover to healthy serving paths if degradation occurs, and exact billing details are returned in the usage.cost field. The model does not support reasoning, vision, or structured outputs. It accepts text input only, has a 5,000-token context window, and is strictly an audio generation endpoint. New accounts receive a $0.50 signup credit to test the pipeline.

Specs

Field	Value
Model ID	`eleven-multilingual-v2`
Best for	Narration, storytelling, brand voiceovers, multilingual content
Context window	5,000
Input modalities	Text
Output modalities	Audio
Tool calling	Yes
Structured outputs	Yes
Prompt caching	Yes
Speed	Medium
Cost band	Balanced
Release stage	Stable

Pricing

	Per 1M tokens
Input	$0.20
Output	$0.60

Use this when

Multilingual Voiceover Generation — Produce localized audio tracks for videos and presentations across 29 supported languages.
Brand Narrative Synthesis — Maintain consistent, brand-safe vocal delivery across marketing and onboarding content.
Long-Form Storytelling Audio — Render expressive narration for audiobooks, podcasts, or interactive media.
Text Translation to Speech — Convert written translations into natural-sounding audio for accessibility or localization workflows.

Not ideal for

Do not use this model for tasks requiring text reasoning, structured JSON output, or ultra-low latency streaming, as it only generates audio and operates at a medium speed tier.

Pick something else when

You need faster audio generation for real-time applications: use eleven-turbo-v2-5 or minimax-speech-turbo.
You require ultra-low latency at a lower cost tier: use eleven-flash-v2-5.
You need to generate background music or sound effects instead of speech: use elevenlabs-music or elevenlabs-sfx.
You need voice cloning or custom voice design capabilities: use minimax-voice-clone or minimax-voice-design.

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-...")

response = client.chat.completions.create(
    model="eleven-multilingual-v2",
    messages=[{"role": "user", "content": "..."}],
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=generate-speech&tools=true&quality_tier=frontier-open"

FAQ

Does this model support prompt caching on Kyma? Yes, repeated prompt prefixes are billed at this model’s cached input rate, reducing costs for batch or iterative generation. Can I use a single API key for this and other models? Yes, Kyma provides one API key for all models and routes requests through an OpenAI-compatible base URL. How does Kyma handle request failures for this model? Every request includes automatic failover, so if a serving path degrades, your request is silently rerouted to a healthy path without manual intervention.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ