Alibaba's TaoLive team open-sources TaoMate-H3, a streaming AV generation runtime on MiniMax H3
bdsqlsz · x · 2026-09-13
Alibaba's TaoLive AIGC team open-sourced TaoMate-H3, a low-latency streaming audio-video generation runtime built on MiniMax H3.
- Three-step LoRA streaming: small chunks generated via three Stage3 denoising intervals, reaching final latent state well before a full H3 request completes
- Joint audio-video generation: speech, sound, and video on one synchronized timeline
- Long-form continuity: clean KV cache and integrated audio guidance preserve visual identity, voice, and motion across prompt boundaries
- Multiple resolutions: 480p/768p/1080p, portrait and landscape
- Single-node inference: 4 or 8 GPUs with TP2 and Ulysses sequence parallelism
Related event: Alibaba's TaoLive open-sources TaoMate-H3 streaming audio-video runtime(2 posts)→
More from Multimodal
- Hooking Mage's MCP into Claude generated narrative image sets in one shot — AIandDesign · 2026-10-31
- At 3 am she stops scrolling and a whole town answers in songs — AI short film — ownubie · 2026-10-31
- Phone-scanned homes let anyone tour houses in a browser — and it costs ~$200 per scan — jonstephens85 · 2026-10-31
- AI image demo: the V1 prompt that later made the arrow fly — charis_ai · 2026-10-31
- NVIDIA open-sources NemotronLabs RE-USE, a 9.6M-param multilingual speech denoiser — NVIDIAAI · 2026-10-31
- AI-generated fake trailer imagines a 1999 Dino Crisis movie with retro VHS feel — noizlab_studio · 2026-10-31