Assemble Bench (robotics C

AI benchmark · Computer Use

Publisher: HUD
Domains: Computer Use Enterprise Workflows
On X: @hud_evals ↗
Tier: C why
Snapshot: October 11, 2026

Open benchmark ↗hud.ai/research

Full listing name: Assemble Bench (robotics, NIST assembly board; CG-DAgger) - 2026-07-31

Latest from HUD on X

@hud_evals ↗

When building RL tasks and evals, it's difficult to distinguish between actual model capability failures and bad RL task design. We're sharing some our task design principles so you can write better tasks before ever evaluating a model on them.
— hud_evals (@hud_evals) Oct 11, 2026
Here in SF this week? We're hosting a happy hour tomorrow evening, Thursday, Oct 8, 5–8 pm at our SF office. If you're into all things data, come say hi! Sign up here:
— hud_evals (@hud_evals) Oct 7, 2026
Most browser evals look clean in a sandbox and fall apart on the live web. We wrote a walkthrough with @browserbase on how to write browser evals that hold up on the live web. Sites update, graders reward faked outputs, and bot detection kill your runs before the agent even starts. If you're benchmarking browser agents, your scores are probably noisier than you think.
— hud_evals (@hud_evals) Sep 28, 2026

What is Assemble Bench (robotics?

Assemble Bench benchmarks robot models on contact-rich assembly tasks.

Robotics mostly
“Assemble Bench: Benchmarking Robot Models on Contact-Rich Assembly”
hud.ai/research ↗

Tier C

Niche: a handful of models scored, or built by an RL vendor.

How much stands behind a tier: usage, who backs it, a confirmed line from its own site or paper, and a clear read on X. High with three of the four, medium with two, low otherwise.

built by HUD
in curated RL directories (2 pts)

How we got to tier C

Adoption: built by HUD. Tier C.
Not yet read on X.
In curated RL directories: 2 points, a floor of C.
Confirmed: a line from its own site or paper places it.

A clear view on X moves a tier one step up or down; most entries don't move. How tiers work.

Training methods

Methods Assemble Bench (robotics uses, offers or builds for, with public evidence: its own website says so, or two independent sites do.

GRPO Group Relative Policy Optimization On its own website
“The fastest way to understand GRPO is to watch a model get better at something you can see. So: take 2048, wrap it as an environment, and train a small model to play it. The whole thing is about 100 lines. The RL parts (environment contract and training…”
hud.ai
“…ready use hud.ai environments for agent evaluation and RL training. hud.ai environments powered Autonomy-10, which evaluated OpenAI Operator at launch. Every evaluation generates trajectory data that feeds directly into GRPO training pipelines. Publishing on hud.ai means frontier labs can run your environment without custom integration, separate hosting, or outreach. No other platform on this list puts your environment where model companies…”
hud.ai

Where Assemble Bench (robotics sits

Its realm in the universe, live. Hover a star for its name; click to open it. Open it in the universe

The galaxy could not load. See every vendor on the board.

More Computer Use AI benchmarks

Computer Use RL environments

Next: browse all benchmarks. Entry from the RL Research daily scrape of public sources, 2026-10-11 snapshot.