FrontierSWE

Mean@5 score on 34 very long (up to 20h) expert-level engineering and research-engineering tasks.

% solvedHigher is betterCounts toward:Weights: Coding ×0.8The test's websiteOfficial page ↗

The best 13 · 13 models tested

Top 13 · best setting per model · 13 models, 15 results (10 independent)

Measured by independent testersIndependent Reported by the makerVendor-reported

FrontierSWE leaderboard
				Thinking levelSetting
1	Kimi K3	Moonshot AI (Kimi)	81.2%	Maxreasoning_effort=max	Maker's own figureVendor-reportedMoonshot AI ↗	Kimi Code	16 Jul 2026
2	GLM-5.3	Z.ai (Zhipu)	78.1%	Maxreasoning_effort=max	Maker's own figureVendor-reportedZ.ai ↗	—	14 Aug 2026
3	GLM-5.2	Z.ai (Zhipu)	74.4%	Maxmax effort	Maker's own figureVendor-reportedZ.ai ↗	—	16 Jun 2026
4	Qwen3.8 Max (0803)	Qwen (Alibaba)	73.5%	Defaultthinking (default reasoning_effort=xhigh; eval setting not stated)	Maker's own figureVendor-reportedQwen (Alibaba) ↗	Claude Code	3 Aug 2026
5	GPT-6 Astra	OpenAI	65.5%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
6	Claude Opus 5.5	Anthropic	62.3%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
7	Gemini 4 Argon	Google (Gemini / DeepMind)	55.0%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
8	Qwen3.7 Max	Qwen (Alibaba)	40.7%	Defaultthinking (setting not stated)	Maker's own figureVendor-reportedQwen (Alibaba) ↗	Claude Code	3 Aug 2026
9	Grok 4.7	SpaceXAI (formerly xAI)	29.5%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
10	Qwen3.8 Max (0902)	Qwen (Alibaba)	15.8%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
11	DeepSeek V4 Flash Vision Exp	DeepSeek	14.8%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
12	Muse Spark 1.2	Meta (Meta Superintelligence Labs, Muse)	12.0%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026
13	Inkling	Thinking Machines Lab	4.1%	Default	Independent testIndependentFrontierSWE ↗	proximus	1 Oct 2026