Daxiao opens ACE-Brain-0.6 API, but evaluation criteria remain inconsistent
What happened
Daxiao Robotics released the ACE-Brain-0.6 API. Developers can call its spatial grounding capability through the ACE Hub platform without deploying the model independently. The company said that the model upgrade completed alongside this API release expanded high-quality spatial grounding training data from about 1.1 million entries to 1.5 million entries, raised the average score on seven public grounding benchmarks to 75.41, and also surpassed GPT-6 Astra in inference speed. However, two previous reports gave differing accounts of the results: an arXiv paper said the model surpassed GPT-6 Astra on four of seven public grounding benchmarks, with an overall average score of 75.41 versus 73.68; a WeChat public account post from Future Robotics Network said it surpassed Astra on six of eight grounding evaluations, with an average of 74.53 versus 73.68 on the seven public benchmarks, and said its self-built real-robot evaluation set RobotGrounding (2,591 questions, 14 task types) scored 94.13, versus 91.10 for Astra. The model is fine-tuned based on Qwen3.5-9B. The team said that under fixed fine-tuning settings, overall grounding performance steadily improved as training data expanded from 260,000 entries to one million entries and covered more grounding scenarios. A report from the Xbotics Embodied Intelligence Lab noted that the API lowers the threshold for calling and preliminary validation, but coordinate transformation, sensor adaptation, planning constraints and execution validation still require systems engineering, and cross-scenario applicability awaits validation in more real tasks.
AI 根据报道生成 · 6 days ago更新
Developments
2 developments
- 10-29 17:47 · 2 reportsDaxiao Robotics releases spatially anchored embodied foundation model ACE-Brain-0.6 and opens APIXbotics具身智能实验室 · 微信公众号：Daxiao Robotics releases spatially anchored embodied foundation model ACE-Brain-0.6 and opens API
- 10-20 13:04 · 2 reportsACE-Brain-0.6 Released: Scaling Spatial Perception of a 9B Embodied Model with Verifiable Grounding SupervisionarXiv cs.RO：ACE-Brain-0.6 Released: Scaling Spatial Perception of a 9B Embodied Model with Verifiable Grounding Supervision
Reporting timeline
Follow the reporting to explore different perspectives.
- Xbotics具身智能实验室 · 微信公众号Daxiao Robotics releases spatially anchored embodied foundation model ACE-Brain-0.6 and opens API
The Daxiao ACE team has released ACE-Brain-0.6, a vision-spatial-anchoring-enhanced embodied foundation model that maps objects, parts, and interaction regions in language instructions to positions in the visual scene, and has opened an API through ACE Hub.
- 未来机器人网 · 微信公众号Daxiao Robotics opens ACE-Brain-0.6 API, bringing spatial grounding capability to ACE Hub
Daxiao Robotics has released the ACE-Brain-0.6 API, opening its spatial grounding capability to external developers through the ACE Hub platform, so they can call it without deploying the model independently. A simultaneous upgrade expanded high-quality spatial grounding training data from about 1.1 million to 1.5 million entries, raised the model's average score on seven public grounding benchmarks to 75.41, and increased inference speed beyond that of GPT-6 Astra.
- 机器人产业应用 · 微信公众号Embodied AI Industry Weekly: ByteDance Expands Robotics R&D Hiring, Unitree Releases H2-X Humanoid Robot, Daxiao ACE-Brain-0.6 Model Outperforms GPT in Evaluations
ByteDance's Seed Model team added ten robotics positions, covering body design, core components, low-level control, application software, and data evaluation, extending from the model layer toward full-robot development. Unitree Technology released the humanoid robot H2-X "Kung Fu Legend Guardian"; a demo video shows a maximum battery life of about 8 hours, the ability to lift a car bare-handed, and a landing from 4 meters high. Daxiao Robotics released the embodied model ACE-Brain-0.6, which surpassed GPT-6Astra in six of eight Grounding evaluations.
- arXiv cs.ROACE-Brain-0.6 Released: Scaling Spatial Perception of a 9B Embodied Model with Verifiable Grounding Supervision
The research team released ACE-Brain-0.6, which uses a unified data pipeline to combine robot manipulation recordings with complementary visual data from multiple sources, automatically annotates them, and converts the result into verifiable grounding supervision; the model was obtained by fine-tuning Qwen3.5-9B on approximately 1.5 million samples. ACE-Brain-0.6 outperforms GPT-6 Astra on four of seven public grounding benchmarks, with an overall average score of 75.41 versus 73.68; under a fixed fine-tuning setup, overall grounding performance improves steadily as the dataset expands to cover more grounding scenarios.
- 未来机器人网 · 微信公众号ACE Robotics releases embodied foundation model ACE-Brain-0.6, surpassing GPT-6 Astra on six of eight Grounding benchmarks
ACE Robotics released ACE-Brain-0.6, a vision-spatial-anchoring-enhanced embodied foundation model focused on Grounding capability. It surpasses GPT-6 Astra on six of eight Grounding benchmarks, with an average score of 74.53 across seven public benchmarks (Astra: 73.68), and scores 94.13 on its self-built real-robot evaluation set RobotGrounding (2,591 questions, 14 task types; Astra: 91.10).
Activity for this event
- Current
- 3
- Peak
- 9Nov 5 22:00
- Past 24 hours
- -49%
The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.