This Free AI Makes Your Characters Actually Act

Open source AI voice & video tools are getting a serious upgrade Two open source releases for voice + lip sync this week — Drama Box from Resemble AI brings emotional, directable text-to-speech, and LTX's new LipDub LoRA lets you swap dialogue in any video while preserving the original performance. Plus, Inception Labs' Mercury 2 — what happens when you replace the brain of an LLM with a diffusion model? (Apparently, Microsoft comes calling.) 🔗 LINKS ► Drama Box (Resemble AI): https://www.resemble.ai/learn/models/dramabox ► Drama Box on Hugging Face: https://huggingface.co/spaces/ResembleAI/Dramabox ► Pinokio(one-click installer): https://pinokio.co/ ► LTX LipDub workflow: https://www.reddit.com/r/comfyui/comments/1tc96q0/lipdub_iclora_from_ltx_23/ ► LTX Director (timeline editor in Comfy):https://github.com/WhatDreamsCost/WhatDreamsCost-ComfyUI ► Mercury 2 Playground (Inception Labs): https://www.inceptionlabs.ai/ ━━━━━━━━━━━━━━━ 🎯 SPONSOR: InVideo Agent One https://invideo.io/i/TheoreticallyMedia ━━━━━━━━━━━━━━━ 00:00 Intro 00:54 Drama Box (Resemble AI) — TTS with Intention 01:54 Demo: Male Talk Show Host 02:57 Why Open Source Wins & Where to Try It 03:33 Installing Drama Box via Pinokio 04:34 Walkthrough: The Drama Box Interface 05:11 Demo: Villain Monologue (Prose Direction) 05:36 Demo: Noir Detective & Voice Reference Tips 06:38 Demo: Cloning My Own Voice (AGI Breaking News) 07:32 Drama Box: The Verdict 08:11 LTX LipDub — Multilingual Joker 09:01 Why LipDub Belongs in App Mode 09:46 LipDub Demo: Sundar Pichai 10:29 LipDub Demo: Renfield at the Tavern 11:04 LTX 13B Requirements & What's Next 11:44 LTX Director — Timeline Editor in Comfy 12:23 Sponsor: InVideo Agent One 15:39 Mercury 2 — The Diffusion LLM Explained 16:43 5x Faster Than Claude Haiku & GPT 5.2 mini 17:26 Mercury 2 Playground Demo 18:21 Microsoft, Inception & Wrap Up