Model comparison
DeepSeek V4.1 Flash vs GPT-4.1
DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 35.9 on the Noometry Index.
Last verified . 26 shared benchmarks.
DeepSeekSummary
- They share 26 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 9 categories and GPT-4.1 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4.1 Flash leads 66.7 to 22.3.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 67.4% for DeepSeek V4.1 Flash and 6% for GPT-4.1.
- DeepSeek V4.1 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 1M.
- DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4.1 Flash | GPT-4.1 | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 52.8 | 35.9 |
| Released | 2026-09-09 | 2025-04-14 |
| Weights | Open | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 393K | 33K |
| Input $ / M tokens | $0.15 | $2 |
| Output $ / M tokens | $0.60 | $8 |
| Results tracked | 37 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 52.9 (#32), GPT-4.1: 34.4 (#238)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Coding | 1506 | 1391 |
| ALE-Bench | 1,092 | 558.1 |
| SWE-bench Verified | — | 48.5% |
| SWE-bench Verified (bash only) | — | 39.6% |
| Aider Polyglot | — | 52.4% |
| LMArena WebDev | 1619 | — |
| SciCode | 51.9% | — |
| WeirdML | — | 39% |
| CadEval | — | 42% |
Agentic & Tool Use GPT-4.1 leads
DeepSeek V4.1 Flash: 31.2 (#69), GPT-4.1: 34.7 (#43)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| APEX-Agents | 39.5% | — |
| Berkeley Function Calling Leaderboard | — | 54% |
| GDP.pdf | 19.8% | — |
Reasoning DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 50.2 (#36), GPT-4.1: 11.7 (#339)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Hard Prompts | 1483 | 1384 |
| DTBench | 89.9% | 68.3% |
| LMCA | 47% | 25.6% |
| Epoch Capabilities Index | 154.9 | 136.78 |
| ARC-AGI-2 | — | 0.4% |
| SimpleBench | — | 27% |
| Kagi LLM Benchmark | — | 52.3% |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | — | 5.5% |
| CritPt | 14.3% | — |
| Chess Puzzles | — | 6% |
| EnigmaEval | — | 2.2% |
| Mystery Game Puzzles | 43% | — |
| Surface Evolver Bench | 46.3% | — |
| ForecastBench | — | 61.5 |
Math DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 66.7 (#25), GPT-4.1: 22.3 (#280)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 67.4% | 6% |
| OTIS Mock AIME 2024-2025 | 98.3% | 38.3% |
| LMArena Math | 1477 | 1370 |
| FrontierMath Tier 4 | 26.8% | — |
| ProofBench | 54% | — |
| Omni-MATH | — | 47.1% |
| MATH Level 5 | — | 83% |
| FrontierMath (Feb 2025 set) | — | 5.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 57.9 (#38), GPT-4.1: 37.1 (#160)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| GPQA Diamond | 89.8% | 66.9% |
| LMArena Expert | 1506 | 1364 |
| Humanity's Last Exam | — | 5.4% |
| SimpleQA Verified | — | 31.1% |
| MMLU-Pro | — | 81.1% |
| Vectara Hallucination Rate | — | 5.6% |
| GPQA (HELM) | — | 65.9% |
Multimodal Too close to call
DeepSeek V4.1 Flash: 39.1 (#61), GPT-4.1: 38.2 (#67)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Vision | 1277 | 1211 |
| GeoBench | — | 72% |
| Furniture Assembly | 34.2% | — |
Multilingual DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 55.0 (#35), GPT-4.1: 49.4 (#133)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Non-English | 1448 | 1370 |
| LMArena Chinese | 1497 | 1382 |
| LMArena French | 1452 | 1382 |
| LMArena German | 1484 | 1381 |
| LMArena Japanese | 1412 | 1319 |
| LMArena Korean | 1452 | 1339 |
| LMArena Russian | 1471 | 1377 |
| LMArena Spanish | 1459 | 1376 |
Instruction Following DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 77.3 (#26), GPT-4.1: 71.3 (#153)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Instruction Following | 1474 | 1367 |
| IFEval | — | 83.8% |
Long Context DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 45.2 (#47), GPT-4.1: 40.0 (#163)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Longer Query | 1475 | 1385 |
| Fiction.LiveBench | — | 63.9% |
Writing & Preference DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 65.4 (#48), GPT-4.1: 57.6 (#125)
| Benchmark | DeepSeek V4.1 Flash | GPT-4.1 |
|---|---|---|
| LMArena Text | 1462 | 1383 |
| LMArena Creative Writing | 1435 | 1363 |
| EQ-Bench Creative Writing | 1540 | 1420 |
| LMArena Multi-Turn | 1457 | 1398 |
| WildBench | — | 85.4% |
Frequently asked questions
Is DeepSeek V4.1 Flash better than GPT-4.1?
DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 35.9 on the Noometry Index.
Which is cheaper, DeepSeek V4.1 Flash or GPT-4.1?
DeepSeek V4.1 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-4.1 lists at $2 and $8.
Is DeepSeek V4.1 Flash or GPT-4.1 better for coding?
DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.9 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 1M.
How many benchmarks do DeepSeek V4.1 Flash and GPT-4.1 share?
26 benchmarks have published results for both models. DeepSeek V4.1 Flash has 37 scored results on Noometry and GPT-4.1 has 52.