← Sinan Lab › Sinan Rankings · Robo
Sinan Rankings · measured boards · updated with the index

Sinan Rankings · Robo

Only fields with a source get ranked: release date, parameter count from the model card, pretraining hours as published, declared target bodies, licence as written. No composite score, no recommendation.

40models indexed12robot bodies17commercial licence, clearly8non-commercial, clearly

Newest releases

By release month on the model card or README

  1. 01InternVLA-A1.5上海人工智能实验室 InternRobotics2026-072.69 B params
  2. 02LingBot-VLA 2.0蚂蚁灵波科技 (Ant Group)2026-076.00 B params
  3. 03Galaxea G0.5星海图(Galaxea)2026-062.00 B params
  4. 04Isaac GR00T N1.7NVIDIA2026-043.00 B params
  5. 05Spirit v1.5千寻智能(Spirit AI)2026-015.40 B params
  6. 06LingBot-VLA 4B蚂蚁灵波(Robbyant)2026-014.20 B params
  7. 07GR00T N1.6 3BNVIDIA2026 (待核实)3.29 B params
  8. 08VLA-0NVIDIA Research2025-103.00 B params

Most pretraining data

Published real-robot / human pretraining hours; only models that state a number

  1. 01LingBot-VLA 2.0蚂蚁灵波科技 (Ant Group)60,000 h
  2. 02LingBot-VLA 4B蚂蚁灵波(Robbyant)20,000 h
  3. 03RDT2-VQ清华大学 TSAIL(thu-ml)10,000 h
  4. 04Galaxea G0.5星海图(Galaxea)500 h

Broadest declared coverage

Number of body forms the vendor declares as targets (declaration, not verified compatibility)

  1. 01LingBot-VLA 2.0humanoid · dual_arm · single_arm · wheeled4 body forms
  2. 02π0.5single_arm · dual_arm · mobile_manipulator3 body forms
  3. 03UniVLAsingle_arm · dual_arm · wheeled3 body forms
  4. 04GR00T N1.5 3Bhumanoid · dual_arm · single_arm3 body forms
  5. 05π0 (pi-zero)single_arm · dual_arm · mobile_manipulator3 body forms
  6. 06RDT-1Bsingle_arm · dual_arm · mobile_manipulator3 body forms
  7. 07InternVLA-A1.5single_arm · dual_arm2 body forms
  8. 08Galaxea G0.5dual_arm_wheeled · single_arm2 body forms

Smallest with a commercial licence

Parameter count from the model card, licence clearly allows commercial use

  1. 01OctoMIT(代码与权重仓库标注)93M B params
  2. 02X-VLAApache-2.0(代码与模型卡)880M B params
  3. 03RDT-1BMIT(代码、权重、数据均为 MIT,README 原文)1.00 B params
  4. 04Isaac GR00T N1.7NVIDIA Open Model License(权重);Apache-2.03.00 B params
  5. 05EO-1MIT(模型卡)3.77 B params
  6. 06SpatialVLAMIT(代码 README 与模型卡均标 MIT)4.03 B params
  7. 07Spirit v1.5代码 MIT(README);模型卡标 apache-2.05.40 B params
  8. 08LingBot-VLA 2.0Apache-2.06.00 B params

Largest open weights

Parameter count from safetensors metadata where available

  1. 01Magma-8BMicrosoft Research8.90 B params
  2. 02RDT2-VQ清华大学 TSAIL(thu-ml)8.29 B params
  3. 03MolmoActAllen Institute for AI(Ai2)8.12 B params
  4. 04Being-H0BeingBeyond(北京)8.11 B params
  5. 05OpenVLAStanford / UC Berkeley / TRI 等(OpenVLA 团队)7.54 B params
  6. 06UniVLAOpenDriveLab(香港大学 等)7.54 B params
  7. 07RynnVLA-001阿里巴巴达摩院7.01 B params
  8. 08LAPAKAIST / 华盛顿大学 等(Latent Action Pretraining)7.00 B params

Most-targeted bodies

How many indexed models declare this body form as a target

  1. 01松灵 Piper 机械臂松灵机器人 · single_arm31 models
  2. 02SO-101开源社区 (LeRobot) · single_arm31 models
  3. 03Franka FR3 / PandaFranka Robotics · single_arm31 models
  4. 04WidowX 250 STrossen Robotics · single_arm31 models
  5. 05Google Robot(Everyday Robots)Google / Everyday Robots · single_arm31 models
  6. 06Franka Emika PandaFranka Robotics · single_arm31 models
  7. 07智元 精灵 G1智元机器人 · dual_arm_wheeled19 models
  8. 08Galaxea R1 Pro星海图(Galaxea) · dual_arm_wheeled19 models

Licence at a glance

Counted from licence text as written; when code and weight licences disagree we count it as unverified

17 commercial OK8 non-commercial15 unverified
Sorted by recorded facts; not a recommendation. Every field links to its evidence on the model page. Data 2026-09-15.
SinanLab · 司南实验室Clear direction, grounded judgement.