# DeepSeek V4.1 Flash vs GPT-5.5

> GPT-5.5 is the stronger model overall, scoring 63.4 to 50.3 on the Noometry Index. DeepSeek V4.1 Flash costs 43× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v4-1-flash-vs-gpt-5-5
- Last updated: 2026-10-10
- Shared benchmarks: 25

## Summary

- They share 25 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 0 categories and GPT-5.5 in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.5 leads 81.5 to 48.4.
- The biggest single-benchmark swing is Surface Evolver Bench: 46.3% for DeepSeek V4.1 Flash and 88.1% for GPT-5.5.
- DeepSeek V4.1 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 1M.
- DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 50.3 | 63.4 |
| Rank | 40 | 8 |
| Context | 1M | 1.05M |
| Input $/M | $0.15 | $5 |
| Output $/M | $0.60 | $30 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek V4.1 Flash: 52.8 (#29)
- GPT-5.5: 58.3 (#13)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena WebDev | 1620 | 1512 |
| SciCode | 51.9% | 56.1% |
| LMArena Coding | 1511 | 1494 |
| ALE-Bench | 1,092 | 1,943 |
| SWE-bench Verified | — | 80.6% |
| DeepSWE | — | 67% |
| FrontierCode | — | 43% |
| GSO | — | 40.2% |
| WeirdML | — | 84.9% |
| MirrorCode | — | 10% |

## Agentic & Tool Use

- DeepSeek V4.1 Flash: —
- GPT-5.5: 50.7 (#6)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| Terminal-Bench | — | 84.7% |
| APEX-Agents | — | 55.1% |
| OSWorld 2.0 | — | 13% |
| Remote Labor Index | — | 6.3% |
| τ²-bench Banking | — | 44.6% |
| DeepResearch Bench | — | 54% |
| PostTrainBench | — | 27.2% |
| ExploitBench | — | 47.4% |
| GBAEval | — | 53.2% |
| GDP.pdf | — | 26% |
| LMArena Search | — | 1242 |
| Vending-Bench 2 | — | 7,524 |

## Reasoning

- DeepSeek V4.1 Flash: 46.2 (#39)
- GPT-5.5: 72.7 (#8)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| NYT Connections (extended) | 89.6% | 96.2% |
| CritPt | 14.3% | 27.1% |
| LMArena Hard Prompts | 1485 | 1489 |
| DTBench | 89.9% | 96% |
| LMCA | 47% | 54.3% |
| Surface Evolver Bench | 46.3% | 88.1% |
| Epoch Capabilities Index | 155.01 | 159.25 |
| ARC-AGI-2 | — | 85% |
| SimpleBench | — | 69% |
| Kagi LLM Benchmark | — | 88.8% |
| ARC-AGI-1 | — | 95% |
| Chess Puzzles | — | 54% |
| EBR-Bench | — | 34.3% |
| Mystery Game Puzzles | — | 56% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | — | 60.6 |

## Math

- DeepSeek V4.1 Flash: 48.4 (#52)
- GPT-5.5: 81.5 (#9)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| ProofBench | 54% | 50% |
| LMArena Math | 1493 | 1486 |
| FrontierMath (Tiers 1-3) | — | 85.3% |
| FrontierMath Tier 4 | — | 72.5% |
| MathArena Final-Answer Competitions | — | 94.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| FrontierMath (Feb 2025 set) | — | 51.7% |
| FrontierMath Erdős | — | 0% |
| FrontierMath Tier 4 (v1) | — | 35.4% |

## Knowledge

- DeepSeek V4.1 Flash: 41.6 (#97)
- GPT-5.5: 64.4 (#14)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena Expert | 1505 | 1508 |
| GPQA Diamond | — | 94% |
| SimpleQA Verified | — | 63% |
| Vectara Hallucination Rate | — | 9.3% |

## Multimodal

- DeepSeek V4.1 Flash: —
- GPT-5.5: 47.1 (#11)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena Vision | — | 1297 |
| Blueprint-Bench 2 | — | 36.2% |
| Furniture Assembly | — | 44.2% |
| LMArena Document | — | 1486 |

## Multilingual

- DeepSeek V4.1 Flash: 55.5 (#29)
- GPT-5.5: 56.5 (#19)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena Non-English | 1452 | 1467 |
| LMArena Chinese | 1506 | 1535 |
| LMArena French | 1448 | 1486 |
| LMArena Russian | 1473 | 1473 |
| LMArena Spanish | 1499 | 1469 |
| LMArena German | — | 1478 |
| LMArena Japanese | — | 1498 |
| LMArena Korean | — | 1461 |

## Instruction Following

- DeepSeek V4.1 Flash: 77.6 (#24)
- GPT-5.5: 77.7 (#18)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena Instruction Following | 1475 | 1479 |

## Long Context

- DeepSeek V4.1 Flash: 45.2 (#46)
- GPT-5.5: 48.3 (#12)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena Longer Query | 1476 | 1484 |
| CL-bench Life | — | 22.2% |

## Writing & Preference

- DeepSeek V4.1 Flash: 65.8 (#43)
- GPT-5.5: 72.8 (#13)

| Benchmark | DeepSeek V4.1 Flash | GPT-5.5 |
|---|---|---|
| LMArena Text | 1465 | 1471 |
| LMArena Creative Writing | 1436 | 1455 |
| EQ-Bench Creative Writing | 1540 | 1844 |
| LMArena Multi-Turn | 1460 | 1477 |
| EQ-Bench 4 | — | 1315 |

## FAQ

### Is DeepSeek V4.1 Flash better than GPT-5.5?

GPT-5.5 is the stronger model overall, scoring 63.4 to 50.3 on the Noometry Index. DeepSeek V4.1 Flash costs 43× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek V4.1 Flash or GPT-5.5?

DeepSeek V4.1 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-5.5 lists at $5 and $30.

### Is DeepSeek V4.1 Flash or GPT-5.5 better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.3 versus 52.8 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 1M.

### How many benchmarks do DeepSeek V4.1 Flash and GPT-5.5 share?

25 benchmarks have published results for both models. DeepSeek V4.1 Flash has 25 scored results on Noometry and GPT-5.5 has 71.