Methodology & benchmarks

A short guide to what the scores mean, where they come from, and the caveats you should keep in mind when picking a model.

Where the numbers come from

Every score on this site is a percentage reported by either the model's vendor in their launch evaluation, or an independent leaderboard. We favor verified, third-party harnesses (Scale AI's SWE-bench Pro, Artificial Analysis, BenchLM) over vendor self-reports where both exist. Where numbers come from different harnesses, treat comparisons as directional, not exact.

The benchmarks

SWE-bench Verified

SWE-Vrf · higher is better

Real GitHub issues, human-verified patches. The standard coding-agent benchmark.

SWE-bench Pro

SWE-Pro · higher is better

2,294 harder real-world GitHub issues. Where the frontier separates now.

Terminal-Bench 2.x

Term-B · higher is better

Agentic command-line workflows — planning, iteration, tool coordination.

GPQA Diamond

GPQA · higher is better

Google-proof graduate-level science Q&A. General reasoning proxy.

MMLU-Pro

MMLU · higher is better

Multi-task language understanding across 120+ academic subjects.

HumanEval+

HumEv · higher is better

Functional code generation from docstrings.

MATH-500

MATH · higher is better

Competition-style math problems.

Humanity's Last Exam

HLE · higher is better

Hardest reasoning benchmark (with tools). ~50% = near frontier.

AIME 2025

AIME · higher is better

American Invitational Mathematics Examination.

OSWorld

OSW · higher is better

Real computer-use tasks across desktop apps.

How to read the scores

  • Top-end clustering.On saturated benchmarks (GPQA Diamond, SWE-bench Verified), differences of 1–2 points are within noise. Don't over-index on them.
  • Hard benchmarks separate the field. SWE-bench Pro, Terminal-Bench 2.x, and HLE are where real gaps show up between frontier models.
  • Pricing is per 1M tokensat standard API list price, input / output. “Blended” uses a 3:1 input-to-output ratio as a rough production-workload estimate.
  • Context window is the vendor-stated maximum. Usable context (where the model still retrieves accurately) is usually lower.

Caveats

  • Vendors sometimes change models silently after launch.
  • Benchmark scaffolding (tools, system prompts) changes scores dramatically.
  • Agentic benchmarks especially reward harness engineering, not just model capability.
  • Always run an eval on your own data before committing to a model in production.

Built by GLM-5.2 · Scores are directional, drawn from public mid-2026 benchmarks.

SWE-bench, GPQA, MMLU, AIME, HLE etc. are trademarks of their respective owners.