Head to head
vs For light moderate heavy
Verdict · measured basis
GPT-5.6 Sol costs $0.0095 per solved task against
Claude Opus 5 at $0.10 — ▼ 10x cheaper .
Over 200 tasks that is $18 a month.
On output token price alone Claude Opus 5 looks 1.2x cheaper. The ranking reverses once you divide by pass rate — the cheaper tokens do not win the task.
Copy link Save scenario Export compare table Planned
Verified pricing, pass rate and cost per solved task for Claude Opus 5 and GPT-5.6 Sol Figure Claude Opus 5 GPT-5.6 Sol Input $/Mtok $5 $5 Output $/Mtok $25 $30 Cached input $/Mtok $0.5 $0.5 Pass rate 100% 100% Cost per solved task $0.10 measured $0.0095 measured Cost per month $19 $2
Cost per month against tasks per month · drag the marker
Monthly cost against tasks per month Log-log lines for GPT-5.6 Sol (measured, $0.0095 per solved task), Claude Opus 5 (measured, $0.10 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks. 10 100 1k 10k 100k $0.1 $1 $10 $100 $1k $10k TASKS / MONTH (log) → 200 tasks GPT-5.6 Sol · MEASURED · $0.0095 per solved task $2 GPT-5.6 Sol MEASURED Claude Opus 5 · MEASURED · $0.10 per solved task $19 Claude Opus 5 MEASURED Monthly cost against tasks per month Log-log lines for GPT-5.6 Sol (measured, $0.0095 per solved task), Claude Opus 5 (measured, $0.10 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks. 10 100 1k 10k 100k $0.1 $1 $10 $100 $1k $10k TASKS / MONTH (log) → 200 tasks GPT-5.6 Sol · MEASURED · $0.0095 per solved task $2 Claude Opus 5 · MEASURED · $0.10 per solved task $19 — GPT-5.6 Sol · MEASURED — Claude Opus 5 · MEASURED Solid measured · dashed modeled · dotted stale. The gap is the verdict.
Solvency Bench (solvency.dev) — first-party measurement · verified 2026-08-21 · prices from each provider's pricing page; measured rows ignore the tier · Methodology · All models at these settings