2026-W32: Real-world issues

Week 32 of 2026 graded 11 coding agents on real-world issues. Gemini 3.8 Flash leads with C-. Grades compare only within the same suite version, and plan usage is each agent's share of its own subscription. Every agent ran the same suite cut headless through its own tools.Measured W32 2026

real-world@v3 - the suite key these grades compare within

Published suite snapshot for 11 graded agents, current through Aug 5, 2026.

Sorted by capability. Plan usage is each agent's share of its own named subscription and is not a cross-agent ranking. See the Real-world issues methodology or each agent's scorecard for full context.

Agent	Grade	Capability	Clears	Task time	Reliability	Plan usage
Gemini 3.8 Flashantigravity-cli	C-	60.0%	tier 4 · partial tier 5	2m 32s	100%	19.6%/1wGoogle AI Pro
Claude Fable 5.1claude-code	C-	60.0%	tier 4 · partial tier 5	2m 49s	100%	122%/5hClaude Max 5x
Claude Opus 5.5claude-code	C-	60.0%	tier 4 · partial tier 5	3m 43s	100%	64%/5hClaude Max 5x
Grok 4.6grok-cli	C-	60.0%	tier 4 · partial tier 5	4m 7s	100%	2.4%/1wSuperGrok Plus
Composer 2.5cursor-cli	C-	56.7%	tier 2 · partial tier 5	1m 57s	100%	1.7%/1mCursor Pro
Gemini 3.1 Proantigravity-cli	D	46.7%	tier 4 · partial tier 5	2m 31s	100%	16.8%/1wGoogle AI Pro
GPT-6 Lunacodex-cli	D	46.7%	tier 4 · partial tier 5	1m 35s	100%	0.7%/1wChatGPT Plus
GPT-6 Solcodex-cli	D	46.7%	tier 4 · partial tier 5	3m 31s	100%	11.4%/1wChatGPT Plus
Claude Sonnet 5claude-code	F	33.3%	tier 3 · partial tier 5	3m 48s	100%	35%/5hClaude Max 5x
Claude Haiku 4.5claude-code	F	30.0%	tier 2 · partial tier 5	3m 9s	100%	17%/5hClaude Max 5x
GPT-5.6 Terracodex-cli	F	27.5%	none	1m 26s	100%	0.9%/1wChatGPT Pro 5x

This page publishes aggregate outcomes only. Task prompts and private checker logic never leave the benchmark. Follow this suite over time on Grades over time.