Head to head
Qwen3.5-Flashvsgpt-oss-20b
For tasks, about a month.
Verdict · measured basis
gpt-oss-20b costs $0.0003 per solved task against
Qwen3.5-Flash at $0.0015 — ▼ 5.4x cheaper.
Over 200 tasks that is $0.24 a month.
On output token price alone gpt-oss-20b looks 2.0x cheaper. The ranking holds, but the margin changes.
Verified pricing, pass rate and cost per solved task for Qwen3.5-Flash and gpt-oss-20b| Figure | Qwen3.5-Flash | gpt-oss-20b |
|---|
| Input $/Mtok | $0.065 | $0.03 |
|---|
| Output $/Mtok | $0.26 | $0.13 |
|---|
| Cached input $/Mtok | $0.065 | $0.03 |
|---|
| Pass rate | 89% | 94% |
|---|
| Cost per solved task | $0.0015measured | $0.0003measured |
|---|
| Cost per month | $0.29 | $0.05 |
|---|
Cost per month against tasks per month · drag the marker
Solid measured · dashed modeled · dotted stale. The gap is the verdict.
Solvency Bench (solvency.dev) — first-party measurement · verified 2026-08-26 · prices from each provider's pricing page; measured rows ignore the tier · Methodology · All models at these settings