Skip to content
SOLVENCY
Request a demo

Account controls are loading.

Head to head

Claude Opus 5vsgpt-oss-20b

For tasks, about a month.

Verdict · measured basis

gpt-oss-20b costs $0.0003 per solved task against Claude Opus 5 at $0.10 — ▼ 365x cheaper. Over 200 tasks that is $19 a month.

On output token price alone gpt-oss-20b looks 192x cheaper. The ranking holds, but the margin changes.

Verified pricing, pass rate and cost per solved task for Claude Opus 5 and gpt-oss-20b
FigureClaude Opus 5gpt-oss-20b
Input $/Mtok$5$0.03
Output $/Mtok$25$0.13
Cached input $/Mtok$0.5$0.03
Pass rate100%94%
Cost per solved task$0.10measured$0.0003measured
Cost per month$19$0.05

Cost per month against tasks per month · drag the marker

Monthly cost against tasks per monthLog-log lines for gpt-oss-20b (measured, $0.0003 per solved task), Claude Opus 5 (measured, $0.10 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks.101001k10k100k$0.01$0.1$1$10$100$1k$10kTASKS / MONTH (log) →200 tasksgpt-oss-20b · MEASURED · $0.0003 per solved task$0.05Claude Opus 5 · MEASURED · $0.10 per solved task$19— gpt-oss-20b · 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