On April 6, 2025, ElevenLabs released Dubbing Studio and Voice Isolation to general availability, graduating both products out of the beta programs they had been running since late 2024. The general availability releases add production-grade reliability SLAs, expanded language coverage, and new API endpoints for programmatic access alongside the existing web-based workflow tools.
Dubbing Studio automates the process of translating and re-recording audio and video content in a target language while preserving the original speaker's voice. The GA release extends supported language pairs to 31 languages and introduces frame-accurate lip-sync export: the system adjusts the timing and pacing of translated speech so that lip movements in the output video align with the audio, a feature that had been in limited preview for select enterprise customers. Speaker diarization, which identifies and separates individual speakers in multi-voice content, now handles up to 34 distinct voices in a single file, up from the previous limit of 13.
The GA API for Dubbing Studio accepts video files up to 2 GB or a URL pointing to publicly accessible video, and returns a project ID that callers poll for status and then use to retrieve the dubbed output. Developers can inject custom glossaries per project to ensure that proper nouns, brand names, and technical terms are translated consistently rather than phonetically adapted. A transcript override endpoint lets callers correct the auto-generated source transcript before dubbing begins, which ElevenLabs identifies as the most common intervention point for professional media localization teams.
Voice Isolation separates a target voice from background music, ambient noise, and other speakers in a mixed audio file. The GA version processes files up to 530 MB, adds support for FLAC and OGG input formats alongside MP3 and WAV, and introduces a stem separation mode that outputs the isolated voice and the separated background track as distinct files — a workflow commonly needed for music production and podcast mastering. Processing is asynchronous through the API, with typical turnaround of 32 to 95 seconds per minute of input audio at current server capacity.
Pricing for Dubbing Studio is consumption-based: $11 per hour of input video dubbed, billed in one-minute increments, with a flat 21% surcharge for lip-sync export. Voice Isolation is priced at $0.11 per minute of input audio. Both products are included in the ElevenLabs Creator, Pro, and Scale plans at a monthly credit allocation, and available as pay-as-you-go credits for the API-only and Starter tiers. Enterprise customers can negotiate volume pricing through a dedicated contract, with SOC 2 Type II compliance documentation available under NDA.
ElevenLabs positioned these releases as infrastructure for the media localization and post-production markets, where dubbing a feature film or episodic series into multiple languages currently requires weeks of studio time and large localization crews. The company cited internal data showing that Dubbing Studio reduces the time to a first complete dubbed draft by 90% compared to traditional workflows, while Voice Isolation's stem separation addresses a long-standing pain point for remix engineers and podcast producers who receive mixed files without access to the original multi-track session.