Another week full of Voice AI updates! ๐๏ธ
More news in live translation ๐ : this time with
@Google's Gemini 3.5 and
@krispHQ releasing Voice Translation v3. Following the release of OpenAI's own model some weeks ago, it looks like everything in this space is accelerating.
In STT updates,
@gladia_io launched a new model showing great results for business speech. On the TTS front,
@MisoLabsAI released Miso One, an updated
@GradiumAI model is significantly reducing errors on emails and numbers, and
@resembleai AI's Chatterbox v3 is improving voice naturalness. On top of that, Inworld cut their Voice AI API prices by ~50%.
In more exotic news,
@ai_coustics released Tyto, a new model that can detect problems in incoming audio in advance. Thereโs also an interesting new opensource tool called NoiseKit, which generates realistic datasets from clean audio for testing and training โค๏ธ
On the platform side,
@livekit launched a new version full of great improvements, plus the release of LiveKit Portalโfeaturing a stack intended for robotics teleoperation on top of their C SDK. Great job in the last weeks by the LiveKit team! ๐
Meanwhile,
@retellai had a massive "Launch Week" featuring Live Monitoring, a Built-in CRM, a Colloquial Model, and Custom Dashboards ๐คฏ The product is looking great and the only close competitor I see these days is ElevenLabs.
Full newsletter with more updates ๐