qvib.pro
RU

SWE-bench and Terminal-Bench

What for: to know what actually measures coding agents — real GitHub issues and terminal tasks instead of multiple-choice quizzes.

open-source профи

SWE-bench Verified — 500 проверенных людьми GitHub-issue; де-факто стандарт оценки кодинг-агентов 2025 checked 2026-06-01

Updated: 02.07.2026

SWE-bench and Terminal-Bench

What it is

This is the family of benchmarks that test agents on real engineering work rather than on a quiz. SWE-bench Verified is 500 human-vetted GitHub issues from popular repos: the agent has to produce a patch that genuinely fixes the bug and passes the tests. Terminal-Bench is ~100 hand-written terminal tasks (build the code, bring up a server, solve an infra problem).

Where it came from

  • 2023 — SWE-bench is released (Princeton). 2024 — OpenAI and Anthropic produce SWE-bench Verified, filtering out broken tasks.
  • 2025 — Terminal-Bench (Stanford + Laude Institute) adds "working in a terminal" as its own axis; SWE-bench Pro, SWE-rebench and others appear.
  • Leaderboards became the public arena where labs compare themselves on every release.

Why it took off

  • Older multiple-choice benchmarks were saturating and said nothing about whether a model can fix real code.
  • "Issue closed / tests pass" is an honest, verifiable metric that mirrors an engineer's job.
  • The numbers became the industry's marketing language: "X% on SWE-bench Verified" is the first thing shown at a model launch.

How to use it today

  1. When comparing models or agents, look at SWE-bench Verified and Terminal-Bench, not at abstract "IQ tests".
  2. Remember the scaffold: the same LLM with different harnesses scores differently — compare the whole setup, not the bare model.
  3. Test your own task on your own code: a public benchmark is a signpost, not a guarantee for your repo.

What to watch out for

  • Contamination: tasks from open repos may have leaked into training — hence the work on inflated metrics (SWE-ABS and similar).
  • A high score doesn't mean "it will handle your legacy": a benchmark isn't your context.
  • The numbers go stale fast — check the date on the leaderboard.

Читать по-русски →