4.1.0

4.1.0 is currently a beta release and must be installed manually — the app will not auto-update to it. Download:

macOS (Apple silicon): https://cdn.feelfish.com/feelfish-releases/v4.1.0-beta.14-20261022T120510/FeelFish-arm64.dmg

macOS (Intel): https://cdn.feelfish.com/feelfish-releases/v4.1.0-beta.14-20261022T120510/FeelFish-x64.dmg

Windows (x64): https://cdn.feelfish.com/feelfish-releases/v4.1.0-beta.14-20261022T120510/FeelFish-Setup-4.1.0-beta.14.exe

🎬 Video preview, image-to-video templates and editing steps are live; video outputs stream to disk
🧰 Media models and custom media APIs in the toolbox, with file-input approval, visual field editing and a media job dashboard
🎙️ New voice-input mode: hold Space to talk, release to insert; recognition runs locally
🔊 New read-aloud bar and auto read-aloud for local, Doubao and online models, plus an audiobook recipe
☁️ Cloud sync now uses a 3 GiB active-file quota, with multipart large-file upload, 30-day recovery and project deletion cleanup
🔌 Custom models support Gemini native protocol; fixed Vertex tool-call 400s and truncated long writes
🔍 Web search now uses DeepSeek, and member fetch uses DeepSeek; 200 search/fetch calls per day with clear quota messages
🧠 Fixed smart-context chapter overwrites and stale deleted-chapter references; added a reset-and-rebuild action
✍️ CSV table editing, credit reset countdown, and aggregated cloud-sync warnings for unsupported file types
🐛 Fixed permanent failures after sub-session compaction, task-list statuses not updating after completion, HarmonyOS keyboard overlap, read-aloud stuck on preparing, quote jumps in separate windows and quota showing 0 B