ByteDanceSeed LiveInterpret
Seed LiveInterpret 2.0
LiveInterpret 2.0 is an end-to-end model for Chinese–English speech interpretation. Its release connects the research model to Volcano Engine’s interpretation service.
Technical guide
The design supports full-duplex conversation and zero-shot voice adaptation. It follows the earlier CLASI system, which produced translated text.Seed LiveInterpret 2.0 release
Practical use
Practical guidance. Assess meaning across a full speaker turn. Names, negation and corrections can arrive late in a sentence, so early spoken output needs a separate accuracy review.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Launch date | |
| Access | API |
| Model identifiers | Doubao Simultaneous Interpretation 2.0 |
| Speech workflow | Speech-to-speech interpretation with voice preservation.Seed LiveInterpret 2.0 release |
Related models
- SeedRealtime
ByteDance · SeedRealtime
Real-time conversation model combining text, audio and visual input.
Launch date not confirmedProvider product