Head to head
vs QW For light moderate heavy
Verdict · measured basis
Qwen3 Coder 30B A3B Instruct costs $0.0001 per solved task against
DeepSeek V3.2 Exp at $0.0001 — ▼ 1.2x cheaper .
Over 200 tasks that is $0.00 a month.
On output token price alone Qwen3 Coder 30B A3B Instruct looks 1.5x 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 DeepSeek V3.2 Exp and Qwen3 Coder 30B A3B Instruct Figure DeepSeek V3.2 Exp Qwen3 Coder 30B A3B Instruct Input $/Mtok $0.27 $0.07 Output $/Mtok $0.41 $0.28 Cached input $/Mtok $0.27 $0.07 Pass rate 81% 61% Cost per solved task $0.0001 measured $0.0001 measured Cost per month $0.02 $0.02
Cost per month against tasks per month · drag the marker
Monthly cost against tasks per month Log-log lines for Qwen3 Coder 30B A3B Instruct (measured, $0.0001 per solved task), DeepSeek V3.2 Exp (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 $10 TASKS / MONTH (log) → 200 tasks Qwen3 Coder 30B A3B Instruct · MEASURED · $0.0001 per solved task $0.02 Qwen3 Coder 30B A3B … MEASURED DeepSeek V3.2 Exp · MEASURED · $0.0001 per solved task $0.02 DeepSeek V3.2 Exp MEASURED Monthly cost against tasks per month Log-log lines for Qwen3 Coder 30B A3B Instruct (measured, $0.0001 per solved task), DeepSeek V3.2 Exp (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 $10 TASKS / MONTH (log) → 200 tasks Qwen3 Coder 30B A3B Instruct · MEASURED · $0.0001 per solved task $0.02 DeepSeek V3.2 Exp · MEASURED · $0.0001 per solved task $0.02 — Qwen3 Coder 30B A3B Instruct · MEASURED — DeepSeek V3.2 Exp · 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