Together AI Launches Voice Finder: Search and Audition 600+ TTS Voices in One Tool
Together AI's new Voice Finder lets developers search, filter, and audition 600+ voices across leading TTS models — by describing the voice they need or uploading an audio sample for similarity matching.

The Problem With TTS Voice Selection
Text-to-speech has exploded in quality over the past two years, but finding the right voice for an application has remained painful. With hundreds of voices spread across a dozen competing TTS models — each with different styles, accents, tones, and emotional ranges — there's been no good way to systematically search for what you actually need.
Together AI's Voice Finder
Together AI has solved this with Voice Finder — a new tool that aggregates 600+ voices across leading TTS models into a single searchable, filterable interface. Developers can find the right voice by:
- Describing it in plain language — "warm, professional female voice with a slight British accent" — and getting matching results
- Uploading an audio sample — for similarity matching when you know roughly what you want but can't articulate it
- Filtering by model, language, style, and other attributes — to narrow down candidates quickly
Why This Matters for AI-Native Apps
Voice is increasingly central to AI products — chatbots, agents, accessibility tools, content generation pipelines, and entertainment applications all need voices that match their brand and use case. Until now, the selection process meant manually previewing voices one by one across multiple provider dashboards. Voice Finder collapses that process into seconds.
Together AI is positioning Voice Finder as infrastructure for AI-native developers — the same audience using their inference API for fast, cheap access to open models. By adding voice discovery and auditioning on top of their TTS API access, they're building a compelling end-to-end developer platform for audio AI.
Getting Started
Voice Finder is available now through Together AI. It's free to explore the voice library, with generation costs tied to your existing Together AI API plan. If you're building anything with voice in 2026, this is a must-bookmark tool.
Enjoyed this?
Get more posts like this delivered to your inbox.
Zero: The Programming Language Built for AI Agents
Next →Serno: The Multi-Agent AI Platform That Builds Expert Teams to Answer Hard Questions
Related Articles
OpenAI Releases Codex Security: An Open-Source CLI and TypeScript SDK to Find, Validate, and Fix Security Vulnerabilities in Your Code
4 min read
Hugging Face Updates MCP Server with Unified Tooling and Sandboxes
4 min read