Head to head
vs GPT-5 is legacy.
For light moderate heavy
No verdict
One cost is modelled and the other is stale . The two bases are never ranked against each other, so no verdict is given. Each figure is shown on its own basis below.
Copy link Save scenario Export compare table Planned
Verified pricing, pass rate and cost per solved task for Gemini 3.1 Pro (preview) and GPT-5 Figure Gemini 3.1 Pro (preview) GPT-5 Input $/Mtok $2 $1.25 Output $/Mtok $12 $10 Cached input $/Mtok $0.2 $0.125 Pass rate 46% 88% Cost per solved task $1.91 modelled $0.74 stale Cost per month $382 $148
Cost per month against tasks per month · drag the marker
Monthly cost against tasks per month Log-log lines for GPT-5 (stale, $0.74 per solved task), Gemini 3.1 Pro (preview) (modelled, $1.91 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks. 10 100 1k 10k 100k $10 $100 $1k $10k $100k TASKS / MONTH (log) → 200 tasks GPT-5 · STALE · $0.74 per solved task $148 GPT-5 STALE Gemini 3.1 Pro (preview) · MODELED · $1.91 per solved task $382 Gemini 3.1 Pro (prev… MODELED Monthly cost against tasks per month Log-log lines for GPT-5 (stale, $0.74 per solved task), Gemini 3.1 Pro (preview) (modelled, $1.91 per solved task) from 10 to 100,000 tasks a month. Marker at 200 tasks. 10 100 1k 10k 100k $10 $100 $1k $10k $100k TASKS / MONTH (log) → 200 tasks GPT-5 · STALE · $0.74 per solved task $148 Gemini 3.1 Pro (preview) · MODELED · $1.91 per solved task $382 — GPT-5 · STALE — Gemini 3.1 Pro (preview) · MODELED Solid measured · dashed modeled · dotted stale. The gap is the verdict.
Source: Scale SEAL leaderboard, SWE-bench Pro · verified 2026-08-21 · prices from each provider's pricing page; measured rows ignore the tier · Methodology · All models at these settings