ElevenLabs

Introducing the Speech Engine Skill Give any agent a voice without leaving your LLM stack. Speech Engine handles speech to text, turn-taking, interruptions, and text to speech, while your server streams the LLM response. Install with > npx skills add elevenlabs/skills

ElevenAgents can be extended with tools to take real action during a conversation. There are three categories: • Client tools • System tools • Integration tools Here's when to use each.

Tools are configured per agent in the dashboard, CLI, or API, and the LLM decides when to call them based on your system prompt. Mix and match across the three categories to build agents that can see, act, route, and respond.

ElevenAgents can be extended with tools to take real action during a conversation. There are three categories: • Client tools • System tools • Integration tools Here's when to use each.

MCP unlocks hundreds of services through one integration. Point your agent at a Zapier MCP server or any custom MCP server you run with HTTP streamable transports.

ElevenAgents can be extended with tools to take real action during a conversation. There are three categories: • Client tools • System tools • Integration tools Here's when to use each.

Integration tools connect your agent to external systems. Three flavors: • Webhooks: call any REST API on your backend • MCP: connect to Model Context Protocol servers • Native integrations: prebuilt connectors for tools like HubSpot

ElevenAgents can be extended with tools to take real action during a conversation. There are three categories: • Client tools • System tools • Integration tools Here's when to use each.

Client tools run directly on the user's device. Use them to trigger UI events, manipulate the DOM, fire notifications, or run any function inside your web or mobile app. If the agent needs to do something visible to the user in real time, this is the right tool.

Text to Speech is the foundation of every voice product you build. Audiobooks, agents, dubbing, accessibility, games, IVR. They all start with turning text into audio that sounds human. Here's how it works in ElevenLabs. Made with ElevenLabs & Hyperframes

Voice Changer transforms any speech into a different voice while preserving the original timing, emotion, and delivery. Send an audio file to our API and get back the same performance in a new voice, ready for character work, localization, or fixing lines in post.

Introducing the Voice Changer Skill Give your agent what it needs to build voice transformation into your app. Convert any voice into another while preserving the original performance, emotion, and timing. Install with > npx skills add elevenlabs/skills

Voice Isolator removes background noise, music, and ambient sound from any recording. Send an audio file to our API and get back a clean speech track, ready for transcription, dubbing, or post-production.

Introducing the Voice Isolator Skill Give your agent what it needs to build voice isolation into your app. Remove background noise and unwanted sounds from any audio. Install with > npx skills add elevenlabs/skills

ElevenLabs Skills have crossed 11k downloads. Once installed, you can use them to build apps with ElevenAgents, Speech to Text, Text to Speech, music, and sound effects. Install them with: > npx skills add elevenlabs/skills

Introducing ElevenAgents React SDK v1.0 A re-architecture of the JavaScript and React SDK to include granular hooks, a unified API across web and React Native, and a stable public API. These changes improve both the developer experience and performance for your end users.

How to upgrade? You can find all the changes in the changelog: github.com/elevenlabs/p... Or we built a skill to help your agent do it for you. Install with: > npx skills add elevenlabs/packages

Introducing ElevenAgents React SDK v1.0 A re-architecture of the JavaScript and React SDK to include granular hooks, a unified API across web and React Native, and a stable public API. These changes improve both the developer experience and performance for your end users.

This release brings six new hooks, each subscribing to an individual slice of conversation state. Components only re-render when the data they consume changes, allowing for better performance.

Introducing Guardrails 2.0 for Agents Control how agents behave in production with protections to keep responses safe, compliant, and reliable.

11k people are using the ElevenLabs MCP server in Claude. It connects to your ElevenLabs account and lets you generate speech, sound effects, and music directly in Claude. All the friction of generating audio with ElevenLabs is gone.

What are Commit Strategies? When you stream audio with ElevenLabs, you get: - Partial transcripts - live interim text. - Committed transcripts - finalized segments after a commit. Commits define structure, latency, and downstream behavior.

Prefer automation. Use VAD. With Voice Activity Detection, we commit automatically when silence is detected. Best for: - Microphone input. - Client-side streaming. You control silence thresholds and sensitivity.

What are Commit Strategies? When you stream audio with ElevenLabs, you get: - Partial transcripts - live interim text. - Committed transcripts - finalized segments after a commit. Commits define structure, latency, and downstream behavior.

Introducing ElevenLabs Skills. With skills, AI coding assistants like Claude Code, Cursor, and OpenCode are better at using our APIs to handle AI audio and agent workflows. Get started: npx skills add elevenlabs/skills

Eleven v3 is out of alpha and ready for commercial use. Since alpha, we've improved stability and accuracy: - Stability - more reliable model and higher user preference scores - Accuracy - 68% fewer errors on numbers, symbols, and technical notation

Here’s an example of how we’ve improved number handling: "+49 170 9876543" Before: "plus forty-nine, one hundred seventy, nine million..." After: "plus four nine, one seven zero, nine eight seven..."