Head to head
QW vs For light moderate heavy
Verdict · measured basis
Qwen3 30B A3B Instruct 2507 costs $0.0000 per solved task against
Mistral Small 3.2 24B at $0.0001 — ▼ 1.2x cheaper .
Over 200 tasks that is $0.00 a month.
On output token price alone Qwen3 30B A3B Instruct 2507 looks 1.0x cheaper. The ranking holds, but the margin changes.
Copy link Save scenario Export compare table Planned
Verified pricing, pass rate and cost per solved task for Qwen3 30B A3B Instruct 2507 and Mistral Small 3.2 24B Figure Qwen3 30B A3B Instruct 2507 Mistral Small 3.2 24B Input $/Mtok $0.048 $0.075 Output $/Mtok $0.193 $0.2 Cached input $/Mtok $0.048 $0.075 Pass rate 69% 69% Cost per solved task $0.0000 measured $0.0001 measured Cost per month $0.01 $0.01
Cost per month against tasks per month · drag the marker
Monthly cost against tasks per month Log-log lines for Qwen3 30B A3B Instruct 2507 (measured, $0.0000 per solved task), Mistral Small 3.2 24B (measured, $0.0001 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks. 10 100 1k 10k 100k $0.001 $0.01 $0.1 $1 TASKS / MONTH (log) → 200 tasks Qwen3 30B A3B Instruct 2507 · MEASURED · $0.0000 per solved task $0.01 Qwen3 30B A3B Instru… MEASURED Mistral Small 3.2 24B · MEASURED · $0.0001 per solved task $0.01 Mistral Small 3.2 24B MEASURED Monthly cost against tasks per month Log-log lines for Qwen3 30B A3B Instruct 2507 (measured, $0.0000 per solved task), Mistral Small 3.2 24B (measured, $0.0001 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks. 10 100 1k 10k 100k $0.001 $0.01 $0.1 $1 TASKS / MONTH (log) → 200 tasks Qwen3 30B A3B Instruct 2507 · MEASURED · $0.0000 per solved task $0.01 Mistral Small 3.2 24B · MEASURED · $0.0001 per solved task $0.01 — Qwen3 30B A3B Instruct 2507 · MEASURED — Mistral Small 3.2 24B · MEASURED 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