Gemini Gets a Voice With Flash TTS Preview

Google Gemini

Latest updates from Google's AI models

🧠 New · Models Gemini 3.1 Flash TTS Preview now available for speech generation Google just dropped a new model that can generate speech in multiple languages with more expressive voices. Think of it like giving Gemini a much better voice — perfect if you're building apps that need to talk to users in a natural way.

Claude Code

What's new from Anthropic's AI coding assistant

🎨 New · Interface Full-screen mode gets flicker-free rendering with new /tui command You can now run /tui fullscreen to get a much smoother visual experience without the annoying screen flickering. It's like switching from an old CRT monitor to a modern display — everything just looks cleaner.

🛠️ New · Commands Separate focus mode control with new /focus command They split up the controls so you can toggle focus mode independently from other view options. Before, Ctrl+O did too many things at once — now it just handles transcript verbosity while /focus handles, well, focus.

🔄 Improved · Session Management Session recap now works even with telemetry disabled If you're using Claude Code through Bedrock, Vertex, or have telemetry turned off, you'll still get session summaries when you return to a conversation. Previously this feature was tied to telemetry, which left some users out in the cold.

🔧 Fixed · Bugs MCP tool calls won't hang when server connections drop · High CPU usage in fullscreen fixed · Session cleanup improvements They squashed several annoying bugs: tools won't get stuck forever if your server connection dies, fullscreen mode won't eat up your CPU when text is selected, and old sessions clean up properly without leaving junk files around.

The AI ChangeLog is generated daily by Demand Signals. We scrape official changelogs, run them through Claude, and publish a plain-English summary so you don't have to read the docs. Subscribe to our blog for daily updates.