Compare / head-to-head
Gemini 3 Pro Previewvs
GPT-5.2
GPT-5.2 leads 9 of 12 shared benchmarks. Gemini 3 Pro Preview is 1.1x cheaper per token. Gemini 3 Pro Preview has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
3 – 9
GPT-5.2 leads
Cheaper per token
Gemini 3 Pro Preview
1.1x cheaper, input + output
Larger context
Gemini 3 Pro Preview
1.0M tokens
Providers
2 providers
Google · OpenAI
Gemini 3 Pro Preview
Google · released 2025-11-18
textvisionaudiovideo
Quality
Benchmark matrix
12 shared · 12 only Gemini 3 Pro Preview · 24 only GPT-5.2| Benchmark | Gemini 3 Pro Preview | GPT-5.2 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 12 | ||||
| SWE-bench Verified | 76.2% ↗ | 80% ↗ | -3.8 pt | GPT-5.2 |
| GPQA Diamond | 91.9% ↗ | 92.4% ↗ | -0.5 pt | GPT-5.2 |
| AIME 2025 | 95% ↗ | 100% ↗ | -5 pt | GPT-5.2 |
| Chess Puzzles (Epoch AI run) | 31% ↗epoch run | 49% ↗epoch run | -18 pt | GPT-5.2 |
| GSO Opt@1 (GSO) | 17.7% ↗ | 26.5% ↗ | -8.8 pt | GPT-5.2 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 91.4% ↗epoch run | 96.1% ↗epoch run | -4.7 pt | GPT-5.2 |
| SAGE (Vals AI) | 47.6% ↗ | 49.3% ↗ | -1.7 pt | GPT-5.2 |
| SimpleBench (SimpleBench) | 76.4% ↗ | 45.8% ↗ | +30.6 pt | Gemini 3 Pro Preview |
| SimpleQA Verified | 72.1% ↗ | 37.1% ↗epoch run | +35 pt | Gemini 3 Pro Preview |
| SWE-bench Verified (Epoch AI run) | 72.9% ↗epoch run | 73.8% ↗epoch run | -0.9 pt | GPT-5.2 |
| Vending-Bench 2 (Andon Labs) | 5478.16 ↗ | 3591.33 ↗ | +1886.8 | Gemini 3 Pro Preview |
| WeirdML (Håvard Tveit Ihle) | 69.9% ↗ | 72.2% ↗ | -2.3 pt | GPT-5.2 |
| Only Gemini 3 Pro Preview reports · 12 | ||||
| Humanity's Last Exam (no tools) | 37.5% ↗ | not reported | — | — |
| AIME 2025 (with code execution) | 100% ↗ | not reported | — | — |
| ARC-AGI-2 (ARC Prize Verified) | 31.1% ↗ | not reported | — | — |
| BALROG (BALROG) | 58.1% ↗ | not reported | — | — |
| CharXiv Reasoning | 81.4% ↗ | not reported | — | — |
| Humanity's Last Exam (with search and code execution) | 45.8% ↗ | not reported | — | — |
| MathArena Apex | 23.4% ↗ | not reported | — | — |
| MMMLU | 91.8% ↗ | not reported | — | — |
| MMMU-Pro | 81% ↗ | not reported | — | — |
| Terminal-Bench 2.0 (Terminus-2 agent) | 54.2% ↗ | not reported | — | — |
| Video-MMMU | 87.6% ↗ | not reported | — | — |
| τ2-bench | 85.4% ↗ | not reported | — | — |
| Only GPT-5.2 reports · 24 | ||||
| AIME 2026 | not reported | 98.3% ↗matharena | — | — |
| LMArena Elo | not reported | 1412.4 ↗ | — | — |
| ARC-AGI-1 (Verified) | not reported | 86.2% ↗ | — | — |
| ARC-AGI-2 (Verified) | not reported | 52.9% ↗ | — | — |
| EBR-bench (Epoch AI run) | not reported | 23% ↗epoch run | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | not reported | 31.7% ↗epoch run | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | not reported | 67.4% ↗epoch run | — | — |
| Furniture Assembly (Epoch AI run) | not reported | 38.3% ↗epoch run | — | — |
| GDPval (wins or ties) | not reported | 70.9% ↗ | — | — |
| LiveBench Agentic Coding (LiveBench) | not reported | 50.3% ↗ | — | — |
| LiveBench Coding (LiveBench) | not reported | 76.1% ↗ | — | — |
| LiveBench Data Analysis (LiveBench) | not reported | 78.2% ↗ | — | — |
| LiveBench Instruction Following (LiveBench) | not reported | 61.8% ↗ | — | — |
| LiveBench Language (LiveBench) | not reported | 79.8% ↗ | — | — |
| LiveBench Mathematics (LiveBench) | not reported | 93.2% ↗ | — | — |
| LiveBench Reasoning (LiveBench) | not reported | 83.2% ↗ | — | — |
| LMArena Vision (LMArena) | not reported | 1242.8 ↗ | — | — |
| LMArena WebDev (LMArena) | not reported | 1415.5 ↗ | — | — |
| Mystery Game Puzzles (Epoch AI run) | not reported | 23% ↗epoch run | — | — |
| SWE-Bench Pro | not reported | 55.6% ↗ | — | — |
| tau2-bench Airline (Sierra) | not reported | 83% ↗ | — | — |
| tau2-bench Banking Knowledge (Sierra) | not reported | 32.2% ↗ | — | — |
| tau2-bench Retail (Sierra) | not reported | 81.6% ↗ | — | — |
| tau2-bench Telecom (Sierra) | not reported | 89.7% ↗ | — | — |
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 Pro Preview minus GPT-5.2 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 Pro Preview | GPT-5.2 | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 400K tokens | Gemini 3 Pro Preview |
| Max output | 66K tokens | 128K tokens | GPT-5.2 |
| Input price / 1M | $2 ↗ | $1.75 ↗ | GPT-5.2 |
| Output price / 1M | $12 ↗ | $14 ↗ | Gemini 3 Pro Preview |
| Cached input / 1M | — | $0.175 ↗ | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $14 | $15.75 | Gemini 3 Pro Preview |
| Modalities | text · vision · audio · video | text · vision | Gemini 3 Pro Preview |
| Released | 2025-11-18 | 2025-12-11 | — |
| Cited benchmark scores | 25 | 45 | — |
Reliability
Provider status
Live from /statusMore matchups
Gemini 3 Pro Preview vs …
Models sharing the most benchmarksMore matchups
GPT-5.2 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Gemini 3 Pro Preview and GPT-5.2 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.