AI Product Updates: New Features and Releases

Gemini 3.8 Flash TTS is out, and in a demo it works out the melody itself and sings a birthday song in a singing style.

Google launched two TTS models, which it calls its most expressive audio generation models yet, available via the Gemini API or in AI Studio.

Recraft launched the V4.1 Flash image generation model, which produces images in about a second, with free credits enough for thousands of images.

TechCrunch reports that creators will be able to edit videos with AI assistance through a conversational chat interface, though the post does not name the platform or a launch date.

LangChain shares Colifran's thread going through the new features and changes in OpenWiki 0.6.0 one by one.

PhAI Labs released ScienceBuddy, a biomedical agent workbench that organizes papers into tables with traceable DOIs, and its Preview version is free to use.

A user shares their experience with Recraft Studio V4.1 Flash, saying its image generation speed is far ahead of other tools.

Motion released an MCP interface that lets Claude Opus 5.5 generate and revise scenes, pacing and visuals for a launch video within the same project.

Google introduced a new text-to-speech experience in AI Studio that costs less than its previous Gemini 3.1 Flash TTS model.

The new text-to-speech models offer 2,000+ production-ready voices, voice replication, 100 languages and a voice design experience, and rank first on Hume AI's voice benchmarks.

Almanac's AI assistant plugs into a clinic's existing tools, covering evidence retrieval, ambient scribing, medical calculators and clinical trial search, with each conversation in its own sandbox and permission asked before it acts.

Google DeepMind launched Gemini 3.8 Flash TTS, which tops a pronunciation robustness benchmark and ranks second on a voice arena, and Gemini 3.8 Flash-Lite TTS, both supporting preset voices and voice replication.