terminalbench

This account is verified. Learn more

@terminalbench

tbench.ai

Joined May 2025

4 Following 875 Followers 29 Posts

Latest posts 5

terminalbench @terminalbench Sep 10

We're hosting a meetup!

Come meet the community behind the benchmark, celebrate recent releases, and join our discussions about the future of agent evaluation.

luma.com/tbench

luma.com Terminal-Bench Meetup @ Laude Lab · Luma Join us at Laude Lab to celebrate the Terminal-Bench community and recent releases. 5:30 // Arrival and food 6:00 // Demos and panel featuring Andy Konwinski…

terminalbench @terminalbench Aug 29

Terminal-Bench 4.0 out now!

Ryan Marten Verified account @ryan_marten Aug 29

We've pushed a version update to the Terminal-Bench dataset and leaderboard.

Terminal-Bench 4.0 calibrates task resources (time, cpu, memory), implements task fixes, and removes saturated tasks.

View quoted post on X

terminalbench @terminalbench Aug 27

Announcing Terminal-Bench-Science!

Steven Dillmann Verified account @StevenDillmann Aug 27

We're releasing Terminal-Bench-Science: a benchmark for evaluating AI agents on research workflows across scientific domains.

An ongoing Stanford-led community effort, built by the team behind Terminal-Bench together with scientific domain experts at research institutions worldwide. v0.1 has 70 tasks. Claude Opus 5 solves only ~30%.

1/n 👇

View quoted post on X

terminalbench @terminalbench Jun 18

Introducing Terminal-Bench Challenges!

A new capability has emerged at the frontier: agents completing large-scale projects autonomously. To test this capability, we felt another flavor of benchmark was needed.

Terminal-Bench Challenges are long-horizon, token-intensive, single-task benchmarks. Today we are releasing our first 3 challenges.

terminalbench @terminalbench May 20

Contribute to Terminal-Bench Science!

Steven Dillmann Verified account @StevenDillmann May 20

📣 Announcing Terminal-Bench Science: benchmarking AI agents on real scientific workflows – now open for task contributions👇

tbench.ai/news/tb-scienc…

@AnthropicAI, @OpenAI, and @GoogleDeepMind use Terminal-Bench to evaluate AI on coding tasks. We're now extending it to scientific workflows.

1/6🧵

View quoted post on X