Frontier BenchAI model benchmarks · Jul 2026
ModelsCompareLabsMethodology
All models
O

GPT-5.5

OpenAI

Best terminal/CLI autonomy and dense academic reasoning. Effort variants up to max.

FrontierProprietary400K contextReleased Apr 2026textvisionaudio-inaudio-outcode

API price · per 1M tokens

$10/$30

input / output

≈ $15.00/M blended

Benchmarks

Bar shows this model's score; tick shows the best score across all tracked models for context.

SWE-bench VerifiedReal GitHub issues, human-verified patches. The standard coding-agent benchmark.
88.7/ best 90.1
SWE-bench Pro2,294 harder real-world GitHub issues. Where the frontier separates now.
58.6/ best 69.2
Terminal-Bench 2.xAgentic command-line workflows — planning, iteration, tool coordination.
82.7/ best 84.0
GPQA DiamondGoogle-proof graduate-level science Q&A. General reasoning proxy.
93.6/ best 94.4
MMLU-ProMulti-task language understanding across 120+ academic subjects.
92.4/ best 93.0
HumanEval+Functional code generation from docstrings.
95.2/ best 96.0
MATH-500Competition-style math problems.
97.1/ best 97.8
Humanity's Last ExamHardest reasoning benchmark (with tools). ~50% = near frontier.
52.2/ best 57.9
AIME 2025American Invitational Mathematics Examination.
95.5/ best 96.5
OSWorldReal computer-use tasks across desktop apps.
78.7/ best 83.4

Other OpenAI models

O

GPT-5.5 Pro

Frontier

≈$26.25/MO

GPT-5.1 Mini

Fast / Cheap

≈$0.70/M

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.