GigaBrain
GigaAIGigaBrain is GigaAI's line of vision-language-action models for robot manipulation. GigaBrain-0.7, the current release, is a 3.5B model whose three-part design combines scene understanding, prediction and action, pretrained on more than 37,000 hours of real-robot, handheld-gripper, egocentric, simulated and world-model-generated data. It ships with training configurations for AgileX, AgiBot and GigaAI's own Maker H01 robots.
GigaWorld, GigaAI's world model, and the Giga-World-Policy checkpoints are separate releases and are not part of this entry.
Openness
3 high confidence- weights
- open (GigaBrain-0.7-3.5B-Base and fine-tunes on the Hub, ungated)
- data
- described (the 37.3k-hour pretraining mix is described, with only sample data published)
- code
- open (pretraining and post-training implementations with per-robot training configs)
- license
- Apache-2.0 (code and weights)
GigaBrain-0.7's weights and its training code are Apache-2.0. The robot data it learned from is described by source and hours, and only a small sample of it is published.
- https://huggingface.co/api/models?author=open-gigaai&limit=100&expand[]=downloads&expand[]=cardData&expand[]=gated&expand[]=createdAt recorded 2026-11-12
GigaBrain-0.7-3.5B-Base (2026-08-24) and its RoboTwin 2.0 and EBench fine-tunes, the newest GigaBrain checkpoints, gated false and tagged apache-2.0.
- https://huggingface.co/open-gigaai/GigaBrain-0.7-3.5B-Base/raw/main/README.md recorded 2026-11-12
GigaBrain-0.7-3.5B-Base card: license apache-2.0; pretraining on "over 37,000 hours of heterogeneous embodied data".
- https://raw.githubusercontent.com/open-gigaai/giga-brain-0/main/LICENSE recorded 2026-11-12
LICENSE: Apache License, Version 2.0.
- https://raw.githubusercontent.com/open-gigaai/giga-brain-0/main/README.md recorded 2026-11-12
README: "Released the GigaBrain-0.7 Code, Model and Sample Data" (2026/08/25) and "the pre-training and post-training implementations"; the 37.3k hours of pretraining data are summarized by source; GigaBrain-0.7-SampleData is "Sample data for running the GigaBrain-0.7 training workflow".
- https://ungh.cc/repos/open-gigaai/giga-brain-0/files/main recorded 2026-11-12
Repository tree: configs/giga_brain_0_from_scratch.py, gb07_pg2_pick_and_place_piper_30k.py, gb07_pg2_push_buttons_h01_30k.py and post-training release configs.
Adoption
2 medium confidenceHugging Face downloads over the trailing 30 days for the five declared GigaBrain checkpoints, most of them for the 0.7 RoboTwin 2.0 fine-tune and base model.
- https://huggingface.co/api/models?author=open-gigaai&limit=100&expand[]=downloads&expand[]=cardData&expand[]=gated&expand[]=createdAt recorded 2026-11-12
GigaBrain-0.7-3.5B-RoboTwin2.0-Clean 6,885, GigaBrain-0.7-3.5B-Base 3,971, GigaBrain-0-3.5B-Base 79, GigaBrain-0.1-3.5B-Base 62 and GigaBrain-0.7-3.5B-EBench 22 downloads in the trailing 30 days (11,019 together).
Capability
4 medium confidenceGigaBrain-0.7 learns from many robot bodies and ships a two-armed policy, level with GR00T. It has no separate reasoning model or whole-body humanoid control of the kind Gemini Robotics pairs with its action model.
- https://huggingface.co/api/models?author=open-gigaai&limit=100&expand[]=downloads&expand[]=cardData&expand[]=gated&expand[]=createdAt recorded 2026-11-12
Hub listing includes GigaBrain-0.7-3.5B-RoboTwin2.0-Clean, the published RoboTwin 2.0 fine-tune.
- https://raw.githubusercontent.com/open-gigaai/giga-brain-0/main/README.md recorded 2026-11-12
README: embodiment IDs 0 to 7 covering AgileX (aloha, mobile_aloha, arx5, ur5), AgiBot G1 and Franka, AgiBot dexterous hands, UMI, EgoDex and the Maker H01; RoboTwin 2.0 "Post-training and simulation evaluation across 50 bimanual tasks".
Verified 2026-11-12