Spain's Speridlabs Open-Sources Idol, a 600M Image-to-3D Model
NielsRogge · x · 2026-11-10
The top trending model on Hugging Face is Idol, an image-to-3D model from Spain-based Speridlabs.
Key specs:
- 600M parameters, pre-trained from scratch
- Under 7.2GB of VRAM required
- Generates 1M splats in 3 seconds
- Apache 2.0 license
The author calls it an impressive open release and links the model, blog and Papers with Code page.
Related event: Image-to-3D Model Idol Tops Hugging Face Trending Chart(2 posts)→
More from Multimodal
- Fable 5.2 Max Turns Istanbul Into an Explorable 3D World for ~$46 — minchoi · 2026-11-11
- Claude Fable 5.2: 10 Wild One-Prompt Demos in Under 24 Hours — minchoi · 2026-11-11
- Fable 5.2 Shows Much More Musical Diversity in Video Spanning Four Decades of C-Pop — dotey · 2026-11-11
- Dev builds full music video with just code using Gemini 4 Carbon — zacharynado · 2026-11-11
- Adobe MAX recap: five Firefly and Creative Cloud AI updates, AI as assistant — rufusd · 2026-11-11
- Behind the scenes of 'Goldfish Mayhem': AI filmmaking meets Cantonese heritage — JunieLauX · 2026-11-11