Head to head
ME vs QW For light moderate heavy
Verdict · measured basis
Qwen2.5 72B Instruct costs $0.0002 per solved task against
Llama 3.2 1B Instruct at $0.0003 — ▼ 2.0x cheaper .
Over 200 tasks that is $0.03 a month.
On output token price alone Llama 3.2 1B Instruct looks 2.0x 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 Llama 3.2 1B Instruct and Qwen2.5 72B Instruct Figure Llama 3.2 1B Instruct Qwen2.5 72B Instruct Input $/Mtok $0.027 $0.36 Output $/Mtok $0.201 $0.4 Cached input $/Mtok $0.027 $0.36 Pass rate 8% 53% Cost per solved task $0.0003 measured $0.0002 measured Cost per month $0.06 $0.03
Cost per month against tasks per month · drag the marker
Monthly cost against tasks per month Log-log lines for Qwen2.5 72B Instruct (measured, $0.0002 per solved task), Llama 3.2 1B Instruct (measured, $0.0003 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 $10 TASKS / MONTH (log) → 200 tasks Qwen2.5 72B Instruct · MEASURED · $0.0002 per solved task $0.03 Qwen2.5 72B Instruct MEASURED Llama 3.2 1B Instruct · MEASURED · $0.0003 per solved task $0.06 Llama 3.2 1B Instruct MEASURED Monthly cost against tasks per month Log-log lines for Qwen2.5 72B Instruct (measured, $0.0002 per solved task), Llama 3.2 1B Instruct (measured, $0.0003 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 $10 TASKS / MONTH (log) → 200 tasks Qwen2.5 72B Instruct · MEASURED · $0.0002 per solved task $0.03 Llama 3.2 1B Instruct · MEASURED · $0.0003 per solved task $0.06 — Qwen2.5 72B Instruct · MEASURED — Llama 3.2 1B Instruct · 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