Electricity Bench

Claude Opus 5 vs GPT-6 Luna

Claude Opus 5 leads overall, C to D. GPT-6 Luna is about 10x cheaper per task. By suite, Claude Opus 5 takes real-world issues (C- to D+) and spec planning (C to C-). Take GPT-6 Luna anyway when you want the side that is cheaper per task and faster on real-world issues.Measured

Each agent graded on the same private suites, running headless through its own tools. Advantages are oriented so a win is a win, whichever way the metric runs.

Verdict: GPT-6 Luna vs Claude Opus 5

Claude Opus 5 leads overall, C to D. By suite: Claude Opus 5 takes real-world issues, C- to D+ (57% vs 50% capability); Claude Opus 5 takes spec planning, C to C- (66% vs 61% capability); Claude Opus 5 takes vibe coding, C+ to F (72% vs 33% capability). Claude Opus 5 used 2% of Claude Max 5x's weekly limit per pass, $0.024 per task at the listed price. GPT-6 Luna used 1% of ChatGPT Plus's weekly limit per pass, $0.002 per task at the listed price; at the ~$100 tier the board defaults to that is an estimated ~0.2% of ChatGPT Pro 5x's weekly limit (est, about $0.002 per task). Per task, GPT-6 Luna is about 10x cheaper ($0.002 vs $0.024). GPT-6 Luna is faster on real-world issues: a median task takes 1m 0s against 1m 56s. Both finished every task on it. Reasons to pick GPT-6 Luna anyway: it is cheaper per task ($0.002 vs $0.024) and it is faster on real-world issues (1m 0s vs 1m 56s median).

Changed since W32: Claude Opus 5: real-world issues usage 3% to 1% of the weekly limit.

Claude Opus 5Clears tier 2 (partial tier 5) at 21.0% of Claude Max 5x per runCOVERALLGPT-6 LunaClears tier 2 (partial tier 5) at 1.0% of ChatGPT Plus per runDOVERALL
real-world@v3C-vsD+
DimensionClaude Opus 5GPT-6 LunaAdvantage
Capability57%50%+7% (Claude Opus 5 leads)
Median task time1m 56s1m 0s56s faster (GPT-6 Luna leads)
Approx. task cost$0.015 / task$0.003 / task$0.012 / task lower (GPT-6 Luna leads)

Claude Opus 5: 21.0% of Claude Max 5x (21.0% / 5h, 1.0% / 1w) per run. Its weekly limit holds about 14.6 of the week's 33.6 5h windows. GPT-6 Luna: 1.0% of ChatGPT Plus (1.0% / 5h, 1.0% / 1w) per run. Its weekly limit holds about 6 of the week's 33.6 5h windows. Each figure is a share of that agent's own plan, so there is no delta to draw.

spec-planning@v1CvsC-
DimensionClaude Opus 5GPT-6 LunaAdvantage
Capability66%61%+5% (Claude Opus 5 leads)
Median task time2m 9s1m 8s1m 1s faster (GPT-6 Luna leads)
Approx. task cost$0.057 / task<$0.011 / task$0.046 / task lower (GPT-6 Luna leads)

Claude Opus 5: 6.0% of Claude Max 5x (6.0% / 5h, 1.0% / 1w) per run. Its weekly limit holds about 14.6 of the week's 33.6 5h windows. GPT-6 Luna: under 1% of ChatGPT Plus per run. Its weekly limit holds about 6 of the week's 33.6 5h windows. Each figure is a share of that agent's own plan, so there is no delta to draw.

vibe-coding@v1C+vsF
DimensionClaude Opus 5GPT-6 LunaAdvantage
Capability72%33%+39% (Claude Opus 5 leads)
Median task time2m 51s1m 37s1m 15s faster (GPT-6 Luna leads)
Approx. task cost$0.011 / tasknot measured-

Claude Opus 5: 1.6% of Claude Max 5x (1.6% / 5h) per good solve. Its weekly limit holds about 14.6 of the week's 33.6 5h windows. GPT-6 Luna: under 1% of ChatGPT Plus (0.4% / 5h) per good solve. Its weekly limit holds about 6 of the week's 33.6 5h windows. Each figure is a share of that agent's own plan, so there is no delta to draw.

Grades and deltas are comparable only within a suite version. See methodology.

Frequently asked questions

Which is better for coding, Claude Opus 5 or GPT-6 Luna?

Both clear tier 2 of 5, but Claude Opus 5 scores higher on capability: 57% versus 50% on real-world@v3, our private suite of real-world coding problems. Latest measurement 2026-09-21.

Which is cheaper per task, Claude Opus 5 or GPT-6 Luna?

Claude Opus 5 used 2% of Claude Max 5x's weekly limit per pass, $0.024 per task at the listed price. GPT-6 Luna used 1% of ChatGPT Plus's weekly limit per pass, $0.002 per task at the listed price; at the ~$100 tier the board defaults to that is an estimated ~0.2% of ChatGPT Pro 5x's weekly limit (est, about $0.002 per task). Per task, GPT-6 Luna is about 10x cheaper ($0.002 vs $0.024).

What do Claude Opus 5 and GPT-6 Luna cost a month?

Claude Opus 5 runs on Claude Max 5x at $100/month (public price as of 2026-07-27). GPT-6 Luna runs on ChatGPT Plus at $20/month (public price as of 2026-07-20).