Cartesia launches Sonic 3.5 and Ink 2, claiming #1 on both real-time TTS and streaming STT leaderboards

Cartesia 发布 Sonic 3.5 与 Ink 2 实时语音模型

Cartesia ships Sonic 3.5 for TTS and Ink 2 for streaming STT as a single real-time voice stack. Sonic 3.5 hits ~82ms time-to-first-audio and ranks #1 on the real-time TTS leaderboard. Ink 2 ranks #1 on Artificial Analysis's streaming STT leaderboard. Cartesia is now the only provider holding both #1 spots. The post does not disclose model parameters, pricing, or release timeline.

Why it matters: Cartesia topping both real-time TTS and STT leaderboards with 82ms latency is a concrete signal worth surfacing. Capped below 85 because voice is a narrower beat and the post doesn't disclose model size or technical details.