# DeepSeek V4.1 Flash vs GPT-5.2

> GPT-5.2 is the stronger model overall, scoring 54.1 to 50.3 on the Noometry Index. DeepSeek V4.1 Flash costs 18× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v4-1-flash-vs-gpt-5-2
- Last updated: 2026-10-10
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 4 categories and GPT-5.2 in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.2 leads 59.2 to 41.6.
- The biggest single-benchmark swing is ProofBench: 54% for DeepSeek V4.1 Flash and 15% for GPT-5.2.
- DeepSeek V4.1 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $1.75 / $14 for GPT-5.2.
- DeepSeek V4.1 Flash accepts more context: 1M tokens versus 400K.
- DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 50.3 | 54.1 |
| Rank | 40 | 31 |
| Context | 1M | 400K |
| Input $/M | $0.15 | $1.75 |
| Output $/M | $0.60 | $14 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek V4.1 Flash: 52.8 (#29)
- GPT-5.2: 52.2 (#32)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena WebDev | 1620 | 1415 |
| LMArena Coding | 1511 | 1447 |
| ALE-Bench | 1,092 | 1,294 |
| SWE-bench Verified | — | 73.8% |
| SWE-bench Verified (bash only) | — | 72.8% |
| SWE-bench Multilingual | — | 66.7% |
| SciCode | 51.9% | — |
| GSO | — | 27.4% |
| WeirdML | — | 72.2% |
| AlgoTune | — | 2.05 |

## Agentic & Tool Use

- DeepSeek V4.1 Flash: —
- GPT-5.2: 40.0 (#24)

| Benchmark | DeepSeek V4.1 Flash | 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

- DeepSeek V4.1 Flash: 46.2 (#39)
- GPT-5.2: 50.1 (#33)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| NYT Connections (extended) | 89.6% | 83.6% |
| LMArena Hard Prompts | 1485 | 1445 |
| DTBench | 89.9% | 90.9% |
| LMCA | 47% | 43.9% |
| Epoch Capabilities Index | 155.01 | 153.45 |
| ARC-AGI-2 | — | 52.9% |
| SimpleBench | — | 45.8% |
| Kagi LLM Benchmark | — | 73.3% |
| ARC-AGI-1 | — | 86.2% |
| CritPt | 14.3% | — |
| Chess Puzzles | — | 49% |
| EnigmaEval | — | 10.4% |
| EBR-Bench | — | 23% |
| Mystery Game Puzzles | — | 23% |
| Surface Evolver Bench | 46.3% | — |
| ForecastBench | — | 60.1 |

## Math

- DeepSeek V4.1 Flash: 48.4 (#52)
- GPT-5.2: 59.8 (#32)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| ProofBench | 54% | 15% |
| LMArena Math | 1493 | 1439 |
| FrontierMath (Tiers 1-3) | — | 67.4% |
| FrontierMath Tier 4 | — | 31.7% |
| MathArena Final-Answer Competitions | — | 72% |
| OTIS Mock AIME 2024-2025 | — | 96.1% |
| FrontierMath (Feb 2025 set) | — | 40.7% |
| FrontierMath Tier 4 (v1) | — | 18.8% |

## Knowledge

- DeepSeek V4.1 Flash: 41.6 (#97)
- GPT-5.2: 59.2 (#29)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena Expert | 1505 | 1445 |
| GPQA Diamond | — | 91.4% |
| Humanity's Last Exam | — | 27.8% |
| SimpleQA Verified | — | 37.1% |
| Vectara Hallucination Rate | — | 8.4% |

## Multimodal

- DeepSeek V4.1 Flash: —
- GPT-5.2: 50.9 (#7)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena Vision | — | 1268 |
| VPCT | — | 84% |
| Furniture Assembly | — | 38.3% |
| LMArena Document | — | 1405 |

## Multilingual

- DeepSeek V4.1 Flash: 55.5 (#29)
- GPT-5.2: 53.6 (#63)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena Non-English | 1452 | 1425 |
| LMArena Chinese | 1506 | 1460 |
| LMArena French | 1448 | 1456 |
| LMArena Russian | 1473 | 1440 |
| LMArena Spanish | 1499 | 1433 |
| LMArena German | — | 1448 |
| LMArena Japanese | — | 1421 |
| LMArena Korean | — | 1390 |

## Instruction Following

- DeepSeek V4.1 Flash: 77.6 (#24)
- GPT-5.2: 74.9 (#83)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena Instruction Following | 1475 | 1417 |

## Long Context

- DeepSeek V4.1 Flash: 45.2 (#46)
- GPT-5.2: 44.0 (#74)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena Longer Query | 1476 | 1428 |
| CL-bench | — | 18.2% |

## Writing & Preference

- DeepSeek V4.1 Flash: 65.8 (#43)
- GPT-5.2: 67.0 (#30)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.2 |
|---|---|---|
| LMArena Text | 1465 | 1439 |
| LMArena Creative Writing | 1436 | 1402 |
| EQ-Bench Creative Writing | 1540 | 1703 |
| LMArena Multi-Turn | 1460 | 1458 |

## FAQ

### Is DeepSeek V4.1 Flash better than GPT-5.2?

GPT-5.2 is the stronger model overall, scoring 54.1 to 50.3 on the Noometry Index. DeepSeek V4.1 Flash costs 18× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek V4.1 Flash or GPT-5.2?

DeepSeek V4.1 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-5.2 lists at $1.75 and $14.

### Is DeepSeek V4.1 Flash or GPT-5.2 better for coding?

They score almost the same on coding (52.8 vs 52.2); test both on your own repository before choosing.

### Which has the bigger context window?

DeepSeek V4.1 Flash does, with 1M tokens against 400K.

### How many benchmarks do DeepSeek V4.1 Flash and GPT-5.2 share?

22 benchmarks have published results for both models. DeepSeek V4.1 Flash has 25 scored results on Noometry and GPT-5.2 has 67.