Electricity Bench

claude-haiku-4-5 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.

claude-haiku-4-5Clears tier 2 (partial tier 5) at 5.0% of Claude Max 20x per runDOVERALLclaude-sonnet-5Clears tier 2 (partial tier 5) at 6.0% of Claude Max 20x per runDOVERALL
real-world@v3CvsD
Dimensionclaude-haiku-4-5claude-sonnet-5Advantage
Capability63%50%+13% (claude-haiku-4-5 leads)
Median task time3m 58s4m 22s24s faster (claude-haiku-4-5 leads)
Approx. task cost$0.005 / task$0.005 / task$0.001 / task lower (claude-haiku-4-5 leads)

claude-haiku-4-5: 5.0% of Claude Max 20x (5.0% / 5h, 1.0% / 1w) per run. claude-sonnet-5: 6.0% of Claude Max 20x (6.0% / 5h, 1.0% / 1w) per run. Each figure is a share of that agent's own plan, so there is no delta to draw.

Work per window: claude-haiku-4-5 2922 vs claude-sonnet-5 2435 suite runs per month (claude-haiku-4-5 leads); per plan-dollar: 14.61 vs 12.18 (claude-haiku-4-5 leads). Work figures compare outputs, so unlike quota shares they are comparable across plans.

spec-planning@v1FvsD
Dimensionclaude-haiku-4-5claude-sonnet-5Advantage
Capability24%39%+15% (claude-sonnet-5 leads)
Median task time3m 4s4m 38s1m 34s faster (claude-haiku-4-5 leads)
Approx. task cost$0.003 / task$0.010 / task$0.007 / task lower (claude-haiku-4-5 leads)

claude-haiku-4-5: 1.0% of Claude Max 20x (1.0% / 5h) per run. claude-sonnet-5: 3.0% of Claude Max 20x (3.0% / 5h) per run. Each figure is a share of that agent's own plan, so there is no delta to draw.

Work per window: claude-haiku-4-5 14610 vs claude-sonnet-5 4870 suite runs per month (claude-haiku-4-5 leads); per plan-dollar: 73.05 vs 24.35 (claude-haiku-4-5 leads). Work figures compare outputs, so unlike quota shares they are comparable across plans.

vibe-coding@v1FvsF
Dimensionclaude-haiku-4-5claude-sonnet-5Advantage
Capability23%34%+11% (claude-sonnet-5 leads)
Median task time4m 18s5m 17s59s faster (claude-haiku-4-5 leads)
Approx. task costnot measured$0.008 / task-

claude-haiku-4-5: 0.0% of Claude Max 20x per good solve. claude-sonnet-5: 0.6% of Claude Max 20x (0.6% / 5h) 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, claude-haiku-4-5 or claude-sonnet-5?

Both clear tier 2 of 5, but claude-haiku-4-5 scores higher on capability: 63% versus 50% on real-world@v3, our private suite of real-world coding problems. Latest measurement 2026-08-18.

Which gives more work per dollar, claude-haiku-4-5 or claude-sonnet-5?

claude-haiku-4-5 delivers more benchmark work per subscription dollar: 14.61 suite runs per plan-dollar versus 12.18 for claude-sonnet-5 (at $200/month priced 2026-08-05, and $200/month priced 2026-08-05). Work figures compare outputs, so unlike raw quota shares they are comparable across plans.

How much do claude-haiku-4-5 and claude-sonnet-5 cost?

claude-haiku-4-5 runs on Claude Max 20x at $200/month (public price as of 2026-08-05). claude-sonnet-5 runs on Claude Max 20x at $200/month (public price as of 2026-08-05).