# GPT-4V vs GPT-5.2

> GPT-5.2 has enough public results to be ranked (#31); GPT-4V does not yet, so treat this comparison as directional.

- Canonical page: https://noometry.com/compare/gpt-4v-vs-gpt-5-2
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Summary

- The widest gap is in multimodal, where GPT-5.2 leads 50.9 to 32.5.

## Snapshot

| | GPT-4V | GPT-5.2 |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 35.8 | 54.1 |
| Rank | — | 31 |
| Context | — | 400K |
| Input $/M | — | $1.75 |
| Output $/M | — | $14 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-4V: —
- GPT-5.2: 52.2 (#32)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| SWE-bench Verified | — | 73.8% |
| SWE-bench Verified (bash only) | — | 72.8% |
| LMArena WebDev | — | 1415 |
| SWE-bench Multilingual | — | 66.7% |
| GSO | — | 27.4% |
| WeirdML | — | 72.2% |
| LMArena Coding | — | 1447 |
| ALE-Bench | — | 1,294 |
| AlgoTune | — | 2.05 |

## Agentic & Tool Use

- GPT-4V: —
- GPT-5.2: 40.0 (#24)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| Terminal-Bench | — | 64.9% |
| Berkeley Function Calling Leaderboard | — | 55.9% |
| GDPval | — | 49.7% |
| Remote Labor Index | — | 2.5% |
| τ²-bench Airline | — | 83% |
| τ²-bench Banking | — | 32.2% |
| τ²-bench Retail | — | 81.6% |
| τ²-bench Telecom | — | 89.7% |
| DeepResearch Bench | — | 41.1% |
| LMArena Search | — | 1207 |
| METR Time Horizons | — | 75.3% |
| Vending-Bench 2 | — | 3,591 |

## Reasoning

- GPT-4V: —
- GPT-5.2: 50.1 (#33)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| ARC-AGI-2 | — | 52.9% |
| SimpleBench | — | 45.8% |
| Kagi LLM Benchmark | — | 73.3% |
| NYT Connections (extended) | — | 83.6% |
| ARC-AGI-1 | — | 86.2% |
| Chess Puzzles | — | 49% |
| EnigmaEval | — | 10.4% |
| EBR-Bench | — | 23% |
| LMArena Hard Prompts | — | 1445 |
| Mystery Game Puzzles | — | 23% |
| DTBench | — | 90.9% |
| LMCA | — | 43.9% |
| Epoch Capabilities Index | — | 153.45 |
| ForecastBench | — | 60.1 |

## Math

- GPT-4V: —
- GPT-5.2: 59.8 (#32)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 67.4% |
| FrontierMath Tier 4 | — | 31.7% |
| MathArena Final-Answer Competitions | — | 72% |
| OTIS Mock AIME 2024-2025 | — | 96.1% |
| ProofBench | — | 15% |
| LMArena Math | — | 1439 |
| FrontierMath (Feb 2025 set) | — | 40.7% |
| FrontierMath Tier 4 (v1) | — | 18.8% |

## Knowledge

- GPT-4V: —
- GPT-5.2: 59.2 (#29)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| GPQA Diamond | — | 91.4% |
| Humanity's Last Exam | — | 27.8% |
| SimpleQA Verified | — | 37.1% |
| Vectara Hallucination Rate | — | 8.4% |
| LMArena Expert | — | 1445 |

## Multimodal

- GPT-4V: 32.5
- GPT-5.2: 50.9 (#7)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| LMArena Vision | — | 1268 |
| Video-MME | 59.9% | — |
| VPCT | — | 84% |
| Furniture Assembly | — | 38.3% |
| LMArena Document | — | 1405 |

## Multilingual

- GPT-4V: —
- GPT-5.2: 53.6 (#63)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| LMArena Non-English | — | 1425 |
| LMArena Chinese | — | 1460 |
| LMArena French | — | 1456 |
| LMArena German | — | 1448 |
| LMArena Japanese | — | 1421 |
| LMArena Korean | — | 1390 |
| LMArena Russian | — | 1440 |
| LMArena Spanish | — | 1433 |

## Instruction Following

- GPT-4V: —
- GPT-5.2: 74.9 (#83)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| LMArena Instruction Following | — | 1417 |

## Long Context

- GPT-4V: —
- GPT-5.2: 44.0 (#74)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| CL-bench | — | 18.2% |
| LMArena Longer Query | — | 1428 |

## Writing & Preference

- GPT-4V: —
- GPT-5.2: 67.0 (#30)

| Benchmark | GPT-4V | GPT-5.2 |
|---|---|---|
| LMArena Text | — | 1439 |
| LMArena Creative Writing | — | 1402 |
| EQ-Bench Creative Writing | — | 1703 |
| LMArena Multi-Turn | — | 1458 |

## FAQ

### Is GPT-4V better than GPT-5.2?

GPT-5.2 has enough public results to be ranked (#31); GPT-4V does not yet, so treat this comparison as directional.

### How many benchmarks do GPT-4V and GPT-5.2 share?

0 benchmarks have published results for both models. GPT-4V has 1 scored results on Noometry and GPT-5.2 has 67.