thereelstack.com · Wed 18 Nov

Wednesday 18 November

Two open TTS models worth knowing for voiceovers (ZONOS3-Mini and Phonon), plus Comfy's official skills plugin for coding agents

ReelStack Lab post

AI-generated summary Zyphra says ZONOS3-Mini is a multilingual voice-cloning model with 100M active parameters and high-fidelity zero-shot voice design capabilities. Gradium offers Phonon as a tiny on-device text-to-speech model with a Rust runtime and Python package support.

Earlier:

OpenMotion goes open source: recreate a reference video with your own brand, plus MOTION.md guides

ReelStack Lab post

AI-generated summary The OpenDesign team launched OpenMotion as an open-source project under the Apache-2.0 license, providing a system where users can recreate reference videos using coding agents. Founder Tom Huang says the project took 14 days and about 470k lines of code, while the team claims recreations reach over 95% fidelity to reference videos without needing Fable 5.2.

Kling Image 4.0: Kuaishou's image model goes native 8K and fits up to 2,000 characters

AI-generated summary Kuaishou unveiled Kling Image 4.0, a new image model featuring native 8K resolution, support for up to 2,000 characters, and a very long prompt limit. Kuaishou says the model is live on the Kling platform and available to businesses and developers through an API.

Also covered bythereelstack.comAITOPAI Ping