Alternatives
Products that do what HunyuanVideo-1.5 does
SOTA video generation on consumer GPUs
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
MiniMax H3▲336MiniMax H3 (aka Hailuo 3) is a multimodal AI video generator. Turn text and images into cinematic clips with native audio and omni-reference — free to start.
Jul 2026 · minimax.io
- 9
- 10
- 11
- 128F
Hi HN! I'm just sharing a project I've been working on during the LLM Efficiency Challenge - you can now finetune Llama with QLoRA 5x faster than Huggingface's original implementation on your own local GPU. Some highlights: 1. Manual autograd engine - hand derived backprop steps. 2. QLoRA / LoRA 80% faster, 50% less memory. 3. All kernels written in OpenAI's Triton language. 4. 0% loss in accuracy - no approximation methods - all exact. 5. No change of hardware necessary. Supports NVIDIA GPUs since 2018+. CUDA 7.5+. 6. Flash Attention support via Xformers. 7. Supports 4bit and 16bit…
2023 · github.com
- 13
- 14
- 15
- 16TT
Hi everyone, I built a UI for the new open-source Hunyuan Motion model to generate 3D animations from text: https://hy-motion.ai It generates BVH files instantly. I'm trying to bridge the gap between "cool AI demo" and "useful game dev tool". Question for 3D devs/animators: If you were to use this in production, what is the single biggest missing feature? 1. Export Pipeline: Auto-conversion to FBX for Unity/Unreal? 2. Motion Fusion: Blending multiple prompts into one long sequence? 3. Rig Variety: Support for non-humanoid skeletons? Feedback is much appreciated.
Jan 2026 · hy-motion.ai
- 17
- 18
- 19
- 20L1
Hey HN, after our 13B release we've been working on a faster version of our open-source video model and we're excited to share it. We started with a 13B base model that already had competitive generation speeds (e.g. 55s for a 5s video on an H100 — faster than any other model out there). But we wanted to push it further to allow everyone to quickly iterate over their video generations. So we built a distilled version focused on speed without sacrificing temporal or spatial coherence. With the Distilled model, you can now generate 5-second 720p videos in about 9.5 seconds on an H100, and…
2025 · lightricks.com
- 21
- 22
- 23
- 24
Ranked by how close each launch is in meaning, then by votes. Refine with a description →