Explore our full suite of AI platforms, data marketplaces, and expert services designed to build, train, fine-tune, and deploy reliable, production-grade AI systems at scale.

Explore our full suite of AI platforms, data marketplaces, and expert services designed to build, train, fine-tune, and deploy reliable, production-grade AI systems at scale.

PRISM evaluation suite · v2.0 · Jun 2026

Benchmarking AI

Driving AI innovation together

Driving AI innovation together

across seven domains

Driving AI innovation together

Driving AI innovation together

The PRISM suite provides rigorous assessments of frontier models across Internationalization, Audio, Vision, Agentic & RL, Physical AI, Healthcare, and AI Safety.

7

7

Domains

14

14

Benchmarks

25K+

25K+

Eval Tasks

50+

50+

Models Evaluated

Abstract image

Physical AI

1 benchmark

Physical & Embodied AI

Physical & Embodied AI

Evaluation of teleoperation, simulation, and policy-transfer for dexterous manipulation on real robot platforms — including the Unitree G1 with BrainCo Revo 2 and Inspire hands across pick-place, fold, pour, and assembly workflows.

Robotic Dexterity Assembly · B4 Dexterous Benchmark

Teleoperation-vs-simulation comparison on the Unitree G1 — 29 dexterous-manipulation datasets · 1,413 episodes · 4 metrics (DMS, GQS, TSR, GAR ratios) and 4 quality gates. 26 / 29 datasets pass all gates. Teleoperation outperforms simulation by up to 80× on Grasp Adaptation Rate (GAR), the decisive metric for real-world policy transfer.

Robotic Dexterity Assembly · B4 Dexterous Benchmark

Teleoperation-vs-simulation comparison on the Unitree G1 — 29 dexterous-manipulation datasets · 1,413 episodes · 4 metrics (DMS, GQS, TSR, GAR ratios) and 4 quality gates. 26 / 29 datasets pass all gates. Teleoperation outperforms simulation by up to 80× on Grasp Adaptation Rate (GAR), the decisive metric for real-world policy transfer.

Connect with Centific

Stay ahead of what’s next

Stay ahead

Updates from the frontier of AI data.

Receive updates on platform improvements, new workflows, evaluation capabilities, data quality enhancements, and best practices for enterprise AI teams.

By proceeding, you agree to our Terms of Use and Privacy Policy