Electricity Bench

gemini-3.1-pro vs claude-sonnet-5

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.

gemini-3.1-proClears tier 3 (partial tier 5) at 8.0% of Google AI Pro per runDOVERALLclaude-sonnet-5Clears tier 3 (partial tier 5) at 9.7% of Claude Max 20x per runDOVERALL
real-world@v3DvsD
Dimensiongemini-3.1-proclaude-sonnet-5Advantage
Capability47%53%+7% (claude-sonnet-5 leads)
Median task time4m 24s5m 6s42s faster (gemini-3.1-pro leads)
Approx. task cost$0.025 / task$0.009 / task$0.016 / task lower (claude-sonnet-5 leads)

gemini-3.1-pro: 8.0% of Google AI Pro (48.1% / 5h, 8.0% / 1w) per run. claude-sonnet-5: about 9.7% of Claude Max 20x (about 9.7% / 5h, 1.0% / 1w) per run (estimated: 2 of 15 cells measured no share of their own). Each figure is a share of that agent's own plan, so there is no delta to draw.

vibe-coding@v1FvsD
Dimensiongemini-3.1-proclaude-sonnet-5Advantage
Capability25%37%+12% (claude-sonnet-5 leads)
Median task time4m 34s5m 21s48s faster (gemini-3.1-pro leads)
Approx. task cost$0.032 / task$0.005 / task$0.027 / task lower (claude-sonnet-5 leads)

gemini-3.1-pro: 0.7% of Google AI Pro (4.2% / 5h, 0.7% / 1w) per good solve. claude-sonnet-5: 0.4% of Claude Max 20x (0.4% / 5h, 0.2% / 1w) per good solve. 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, gemini-3.1-pro or claude-sonnet-5?

Both clear tier 3 of 5, but claude-sonnet-5 scores higher on capability: 53% versus 47% on real-world@v3, our private suite of real-world coding problems. Latest measurement 2026-08-10.