Compare / head-to-head
Gemini 3.1 Pro PreviewvsGemini 3.5 Flash
Gemini 3.1 Pro Preview leads 16 of 29 shared benchmarks. Gemini 3.5 Flash is 1.3x cheaper per token. Both offer a 1.0M-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
16 – 11
Gemini 3.1 Pro Preview leads · 2 tied
Cheaper per token
Gemini 3.5 Flash
1.3x cheaper, input + output
Larger context
Tie
both 1.0M tokens
Providers
Google
same provider
Gemini 3.1 Pro Preview
Google · released 2026-02-19
textvisionaudio
Gemini 3.5 Flash
Google · released 2026-05-19
textvisionaudio
Quality
Benchmark matrix
29 shared · 7 only Gemini 3.1 Pro Preview · 7 only Gemini 3.5 Flash| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.5 Flash | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 29 | ||||
| GPQA Diamond | 94.3% ↗ | 92.8% ↗epoch run | +1.5 pt | Gemini 3.1 Pro Preview |
| AIME 2026 | 98.3% ↗matharena ⚠ | 95% ↗matharena ⚠ | +3.3 pt | Gemini 3.1 Pro Preview |
| LMArena Elo | 1480.1 ↗ | 1477.4 ↗ | +2.7 | Gemini 3.1 Pro Preview |
| APEX-Agents (Mercor) | 35.3% ↗ | 27.5% ↗ | +7.8 pt | Gemini 3.1 Pro Preview |
| ARC-AGI-2 | 77.1% ↗ | 72.1% ↗ | +5 pt | Gemini 3.1 Pro Preview |
| Chess Puzzles (Epoch AI run) | 55% ↗epoch run | 50% ↗epoch run | +5 pt | Gemini 3.1 Pro Preview |
| DeepSWE v1.1 (Datacurve) | 11.7% ↗ | 36.1% ↗ | -24.4 pt | Gemini 3.5 Flash |
| EBR-bench (Epoch AI run) | 14.3% ↗epoch run | 4.8% ↗epoch run | +9.5 pt | Gemini 3.1 Pro Preview |
| FrontierMath Tier 4 v2 (Epoch AI run) | 26.8% ↗epoch run | 26.8% ↗epoch run | 0 | Tie |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 59.6% ↗epoch run | 62.8% ↗epoch run | -3.2 pt | Gemini 3.5 Flash |
| LiveBench Agentic Coding (LiveBench) | 44.1% ↗ | 49% ↗ | -4.9 pt | Gemini 3.5 Flash |
| LiveBench Coding (LiveBench) | 76.5% ↗ | 78.2% ↗ | -1.7 pt | Gemini 3.5 Flash |
| LiveBench Data Analysis (LiveBench) | 78.5% ↗ | 64.9% ↗ | +13.6 pt | Gemini 3.1 Pro Preview |
| LiveBench Instruction Following (LiveBench) | 79.1% ↗ | 75.6% ↗ | +3.5 pt | Gemini 3.1 Pro Preview |
| LiveBench Language (LiveBench) | 85.4% ↗ | 84.6% ↗ | +0.8 pt | Gemini 3.1 Pro Preview |
| LiveBench Mathematics (LiveBench) | 91% ↗ | 88.2% ↗ | +2.8 pt | Gemini 3.1 Pro Preview |
| LiveBench Reasoning (LiveBench) | 84% ↗ | 82% ↗ | +2 pt | Gemini 3.1 Pro Preview |
| LMArena Vision (LMArena) | 1279.3 ↗ | 1284.4 ↗ | -5.1 | Gemini 3.5 Flash |
| LMArena WebDev (LMArena) | 1446.2 ↗ | 1499 ↗ | -52.8 | Gemini 3.5 Flash |
| Mystery Game Puzzles (Epoch AI run) | 34% ↗epoch run | 32% ↗epoch run | +2 pt | Gemini 3.1 Pro Preview |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 95.6% ↗epoch run | 95.6% ↗epoch run | 0 | Tie |
| SAGE (Vals AI) | 48.7% ↗ | 49.9% ↗ | -1.2 pt | Gemini 3.5 Flash |
| SimpleBench (SimpleBench) | 79.6% ↗ | 76.7% ↗ | +2.9 pt | Gemini 3.1 Pro Preview |
| SimpleQA Verified | 73.5% ↗epoch run | 66.2% ↗epoch run | +7.3 pt | Gemini 3.1 Pro Preview |
| SWE-bench Verified (Epoch AI run) | 75.6% ↗epoch run | 79.3% ↗epoch run | -3.7 pt | Gemini 3.5 Flash |
| Terminal-Bench 4.0 (Vals AI) | 2.5% ↗ | 6.1% ↗ | -3.6 pt | Gemini 3.5 Flash |
| Toolathlon-Verified (HKUST) | 61.1% ↗ | 67.3% ↗ | -6.2 pt | Gemini 3.5 Flash |
| Vending-Bench 2 (Andon Labs) | 3774.25 ↗ | 5396.42 ↗ | -1622.2 | Gemini 3.5 Flash |
| WeirdML (Håvard Tveit Ihle) | 72.1% ↗ | 62.6% ↗ | +9.5 pt | Gemini 3.1 Pro Preview |
| Only Gemini 3.1 Pro Preview reports · 7 | ||||
| BALROG (BALROG) | 57% ↗ | not reported | — | — |
| Furniture Assembly (Epoch AI run) | 26.7% ↗epoch run | not reported | — | — |
| GSO Opt@1 (GSO) | 21.6% ↗ | not reported | — | — |
| LMArena Agent (LMArena) | -0.0771 ↗ | not reported | — | — |
| MirrorCode (Epoch AI run) | 8.9% ↗epoch run | not reported | — | — |
| SWE-Bench Pro | 54.2% ↗ | not reported | — | — |
| tau2-bench Banking Knowledge (Sierra) | 26% ↗ | not reported | — | — |
| Only Gemini 3.5 Flash reports · 7 | ||||
| GDPVal-AA | not reported | 1656 ↗ | — | — |
| Humanity's Last Exam (full set, text + MM) | not reported | 40.2% ↗ | — | — |
| MCP Atlas | not reported | 83.6% ↗ | — | — |
| MMMU-Pro | not reported | 83.6% ↗ | — | — |
| OSWorld-Verified | not reported | 78.4% ↗ | — | — |
| SWE-Bench Pro (Public) | not reported | 55.1% ↗ | — | — |
| Terminal-bench 2.1 | not reported | 76.2% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Gemini 3.1 Pro Preview minus Gemini 3.5 Flash in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Gemini 3.1 Pro Preview | Gemini 3.5 Flash | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 1.0M tokens | Tie |
| Max output | 66K tokens | 66K tokens | Tie |
| Input price / 1M | $2 ↗ | $1.5 ↗ | Gemini 3.5 Flash |
| Output price / 1M | $12 ↗ | $9 ↗ | Gemini 3.5 Flash |
| Cached input / 1M | $0.2 ↗ | $0.15 ↗ | Gemini 3.5 Flash |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $14 | $10.5 | Gemini 3.5 Flash |
| Modalities | text · vision · audio | text · vision · audio | Tie |
| Released | 2026-02-19 | 2026-05-19 | — |
| Cited benchmark scores | 43 | 42 | — |
Reliability
Google status
Live from /statusMore matchups
Gemini 3.1 Pro Preview vs …
Models sharing the most benchmarksMore matchups
Gemini 3.5 Flash vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Gemini 3.1 Pro Preview and Gemini 3.5 Flash on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.