# Gemini 3 Pro vs GPT-5.2-Codex

> Side-by-side benchmark comparison for Gemini 3 Pro and GPT-5.2-Codex across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/gemini-3-pro-vs-gpt-5-2-codex
- Last updated: October 17, 2026

- Shared sourced benchmarks: 3
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

Evidence is too thin for a benchmark verdict: Gemini 3 Pro and GPT-5.2-Codex share only 3 sourced benchmark rows.

## Summary

- Only 3 shared sourced benchmark rows are available. That is not enough for a broad capability verdict, so this mirror does not name an overall winner.
- Gemini 3 Pro has the lower published output-token price.
- Gemini 3 Pro has the larger published context window at 2M.

## Model Snapshot

| Property | Gemini 3 Pro | GPT-5.2-Codex |
|----------|----------|----------|
| Creator | Google | OpenAI |
| Type | Proprietary | Proprietary |
| Reasoning | Non-Reasoning | Reasoning |
| Context | 2M | 400K |
| Independent public score (not head-to-head) | 61.18 | 55.3 |
| Benchmarks Covered | 27 | 5 |
| Pricing (input/output) | $2.00 / $12.00 | $1.75 / $14.00 |

## Category Breakdown

### Agentic

- Winner: Insufficient shared evidence
- Gemini 3 Pro public-lane score: Coming soon
- GPT-5.2-Codex public-lane score: 40.8 (Estimated · #64/128)

| Benchmark | Gemini 3 Pro | GPT-5.2-Codex | Winner |
|-----------|-----------|-----------|--------|
| Gert Labs | 63.23% | 51.79% | Gemini 3 Pro |
| JobBench | 11.4% | 26.0% | GPT-5.2-Codex |

### Coding

- Winner: Insufficient shared evidence
- Gemini 3 Pro public-lane score: 43 (Estimated · #59/150)
- GPT-5.2-Codex public-lane score: 43.7 (Supported · #56/150)

| Benchmark | Gemini 3 Pro | GPT-5.2-Codex | Winner |
|-----------|-----------|-----------|--------|
| Vibe Code Bench | 14.30% | 37.91% | GPT-5.2-Codex |
| LiveCodeBench (Vals) | Coming soon | 88.0% | Coming soon |
| SWE-bench (Vals) | Coming soon | 72.4% | Coming soon |

### Reasoning

- Winner: Insufficient shared evidence
- Gemini 3 Pro public-lane score: 50.1 (Unranked · 4 rankable rows)
- GPT-5.2-Codex public-lane score: 82.4 (Unranked · 2 rankable rows)

| Benchmark | Gemini 3 Pro | GPT-5.2-Codex | Winner |
|-----------|-----------|-----------|--------|
| ARC-AGI-2 | 31.1% | Coming soon | Coming soon |
| ARC-AGI-1 | 75.00% | Coming soon | Coming soon |

### Multimodal & Grounded

- Winner: Insufficient shared evidence
- Gemini 3 Pro public-lane score: 80 (#18/54)
- GPT-5.2-Codex public-lane score: 76.9 (Unranked · 1 rankable row)

| Benchmark | Gemini 3 Pro | GPT-5.2-Codex | Winner |
|-----------|-----------|-----------|--------|
| MMMU | 87.2% | Coming soon | Coming soon |
| MMMU-Pro | 81% | Coming soon | Coming soon |
| RealWorldQA | 83.3% | Coming soon | Coming soon |
| OmniDocBench 1.5 | 88.5% | Coming soon | Coming soon |
| Video-MME (with subtitle) | 88.4% | Coming soon | Coming soon |
| Video-MME (w/o subtitle) | 87.7% | Coming soon | Coming soon |
| MathVision | 86.6% | Coming soon | Coming soon |
| We-Math | 86.9% | Coming soon | Coming soon |
| DynaMath | 85.1% | Coming soon | Coming soon |
| SimpleVQA | 73.2% | Coming soon | Coming soon |
| MMLongBench-Doc | 60.5% | Coming soon | Coming soon |
| CC-OCR | 79.0% | Coming soon | Coming soon |
| AI2D_TEST | 94.1% | Coming soon | Coming soon |
| CountBench | 97.3% | Coming soon | Coming soon |
| RefCOCO (avg) | 84.1% | Coming soon | Coming soon |
| ODINW13 | 46.3% | Coming soon | Coming soon |
| ERQA | 70.5% | Coming soon | Coming soon |
| VideoMMMU | 87.6% | Coming soon | Coming soon |
| MLVU (M-Avg) | 83.0% | Coming soon | Coming soon |
| ScreenSpot Pro | 72.7% | Coming soon | Coming soon |
| CharXiv | 81.4% | Coming soon | Coming soon |
| V* | 88.0% | Coming soon | Coming soon |

### Knowledge

- Winner: Insufficient shared evidence
- Gemini 3 Pro public-lane score: 60.8 (Estimated · #47/184)
- GPT-5.2-Codex public-lane score: 56.4 (Estimated · #54/184)

### Instruction Following

- Winner: Insufficient shared evidence
- Gemini 3 Pro public-lane score: 86.1 (#45/148)
- GPT-5.2-Codex public-lane score: 93.8 (#4/148)

## FAQ

### Is there enough evidence to pick between Gemini 3 Pro and GPT-5.2-Codex?

Not yet. The models share only 3 sourced benchmark rows, so this mirror reports the available measurements and metadata without naming a broad winner.

### What can I compare safely on this page?

Published pricing, context windows, model type, and individual shared benchmark rows are directly comparable. Independent overall scores and non-overlapping benchmark coverage are labeled separately and should not be read as a controlled head-to-head result.

## Related Comparisons

- [Gemini 3 Pro vs Gemini 3 Pro Deep Think](/compare/gemini-3-pro-vs-gemini-3-pro-deep-think)
- [GPT-5.2-Codex vs Gemini 3 Pro Deep Think](/compare/gemini-3-pro-deep-think-vs-gpt-5-2-codex)
- [Gemini 3 Pro vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-gemini-3-pro)
- [GPT-5.2-Codex vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-gpt-5-2-codex)

## Explore More

- [Gemini 3 Pro profile](/models/gemini-3-pro)
- [GPT-5.2-Codex profile](/models/gpt-5-2-codex)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)