DeepSeek V4 Pro 0423 vs GPT-5.1
Updated October 21, 2026. 63 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick GPT-5.1 for coding: it scores 53.3 on the Vibe Coding Index against 35.3 for DeepSeek V4 Pro 0423.
Pick DeepSeek V4 Pro 0423 when output tokens cost 5.2x less, or when a solved task costs $0.57 against $1.02.
- Shared results
- 63
- DeepSeek V4 Pro 0423 only
- 20
- GPT-5.1 only
- 37
- Like-for-like categories
- 8 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, four public coding benchmarks
GPT-5.1
GPT-5.1 scores 53.3 against 35.3.
1 and 2 benchmarks behind the scores
Intelligence
AA Intelligence Index
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 leads, 30.4 against 24.7.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only DeepSeek V4 Pro 0423 has a AA Agentic Index result, 26.3.
Directional only
Reasoning
GPQA Diamond
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 leads, 89.4% against 86.6%.
Same board, both measured
Math
AIME
No call
Only GPT-5.1 has a AIME result, 93.3%.
Directional only
Knowledge
MMLU Pro
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 leads, 87.3% against 86.4%.
Same board, both measured
Web design
Design Arena: websites
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 leads, 1235 against 1187.
Same board, both measured
Long documents
Listed context window
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 takes 1.05M tokens against 400K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 $0.0019, GPT-5.1 $0.0063 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
DeepSeek V4 Pro 0423
DeepSeek V4 Pro 0423 $0.57, GPT-5.1 $1.02.
Published attempt cost and pass rate
At a glance
- Coding index
- DeepSeek V4 Pro 0423 35.3 from 1 benchmark, GPT-5.1 53.3 from 2 benchmarks.
- Widest category
- Agentic work, 11.3 points apart on average over 1 shared benchmark.
- Widest single gap
- Vibe Code Bench (Vals AI): DeepSeek V4 Pro 0423 49.9% against GPT-5.1 24.6%, 25.3 points apart on a 0 to 100 scale.
- Output price
- DeepSeek V4 Pro 0423 $1.9105 against GPT-5.1 $10.00 per million output tokens, 5.2x apart.
- Context window
- DeepSeek V4 Pro 0423 takes 1.05M tokens, GPT-5.1 400K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | DeepSeek V4 Pro 0423 | Bars | GPT-5.1 | Ties |
|---|---|---|---|---|
| Coding | 9 | 1 | ||
| Intelligence | 1 | 1 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 4 | 3 | ||
| Math | 2 | 1 | ||
| Knowledge | 5 | 6 | ||
| Web design | 6 | 2 | ||
| General | 16 | 5 | ||
| All shared boards | 44 | 19 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index91
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| SWE-bench Verified | Not measured | 66.0% | Only GPT-5.1 measured |
| LMArena WebDev | 1447 | 1396 | DeepSeek V4 Pro 0423 |
| AA Coding Index | 59.4 | 49.4 | DeepSeek V4 Pro 0423 |
| LiveCodeBench (Vals AI) | 87.5% | 86.5% | DeepSeek V4 Pro 0423 |
| SWE-bench Verified (Vals AI) | 77.4% | 69.8% | DeepSeek V4 Pro 0423 |
| Terminal-Bench 4 (Vals AI) | 11.1% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Vibe Code Bench (Vals AI) | 49.9% | 24.6% | DeepSeek V4 Pro 0423 |
| Code migration (Vals AI) | 26.2% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| ProgramBench (Vals AI) | 0.0% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Terminal-Bench 2.1 (Vals AI) | 50.2% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| SkillsBench (Vals AI) | 51.3% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 86.8% | Only GPT-5.1 measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 45.5% | Only GPT-5.1 measured |
| Terminal-Bench | Not measured | 47.6% | Only GPT-5.1 measured |
| FrontierCode | 17.6% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| GSO-Bench | Not measured | 13.7% | Only GPT-5.1 measured |
| WeirdML | 48.9% | 60.8% | GPT-5.1 |
| ALE-Bench | 1006.1 | Not measured | Only DeepSeek V4 Pro 0423 measured |
| LMArena Text: coding | 1470 | 1453 | DeepSeek V4 Pro 0423 |
| LMArena Text: software and IT | 1469 | 1454 | DeepSeek V4 Pro 0423 |
| LMArena WebDev: React | 1441 | 1307 | DeepSeek V4 Pro 0423 |
| LMArena WebDev: HTML | 1435 | 1405 | DeepSeek V4 Pro 0423 |
| LMArena WebDev: from an image | Not measured | 1421 | Only GPT-5.1 measured |
| Neither model has a result on the other 15 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| AA Intelligence Index | 30.4 | 24.7 | DeepSeek V4 Pro 0423 |
| Epoch Capabilities Index | 149.1 | 149.8 | GPT-5.1 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| AA Agentic Index | 26.3 | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Design Arena: agents building web apps | 1000 | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Design Arena: agents building full-stack apps | 948 | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Design Arena: agents building Godot games | 1059 | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Finance Agent (Vals AI) | 44.1% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Tax Agent Bench (Vals AI) | 25.2% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Legal Research Bench (Vals AI) | 23.1% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| Excel Modeling Benchmark (Vals AI) | 51.6% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| τ²-Bench Telecom | Not measured | 81.9% | Only GPT-5.1 measured |
| DeepResearch Bench | Not measured | 42.8% | Only GPT-5.1 measured |
| Vending-Bench 2 | $3,285 | $1,473 | DeepSeek V4 Pro 0423 |
| Neither model has a result on the other 28 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index43
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| GPQA Diamond | 89.4% | 86.6% | DeepSeek V4 Pro 0423 |
| MMMU Pro (Vals AI) | Not measured | 83.2% | Only GPT-5.1 measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 87.3% | Only GPT-5.1 measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 28.5% | Only GPT-5.1 measured |
| GPQA Diamond (Epoch AI run) | 90.9% | 87.6% | DeepSeek V4 Pro 0423 |
| ARC-AGI | Not measured | 72.8% | Only GPT-5.1 measured |
| ARC-AGI-2 | Not measured | 17.6% | Only GPT-5.1 measured |
| SimpleBench | Not measured | 53.2% | Only GPT-5.1 measured |
| Chess puzzles | 19.0% | 42.0% | GPT-5.1 |
| Mystery game puzzles | 17.0% | 19.0% | GPT-5.1 |
| VPCT | Not measured | 58.7% | Only GPT-5.1 measured |
| LMArena Text: hard prompts | 1461 | 1457 | DeepSeek V4 Pro 0423 |
| LMArena Text: expert | 1463 | 1470 | GPT-5.1 |
| LMArena Text: hard prompts in English | 1466 | 1462 | DeepSeek V4 Pro 0423 |
| LMArena Vision: diagrams | Not measured | 1264 | Only GPT-5.1 measured |
| LMArena Vision: homework | Not measured | 1272 | Only GPT-5.1 measured |
| Neither model has a result on the other 5 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index21
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| AIME | Not measured | 93.3% | Only GPT-5.1 measured |
| MGSM (Vals AI) | Not measured | 93.0% | Only GPT-5.1 measured |
| AA Math Index | Not measured | 94.0 | Only GPT-5.1 measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 94.0% | Only GPT-5.1 measured |
| OTIS Mock AIME | 96.7% | 88.6% | DeepSeek V4 Pro 0423 |
| FrontierMath, tiers 1 to 3 | 45.3% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| FrontierMath, tier 4 | 2.4% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| ProofBench | 16.0% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| LMArena Text: math | 1443 | 1446 | GPT-5.1 |
| LMArena Text: mathematical work | 1455 | 1449 | DeepSeek V4 Pro 0423 |
| Neither model has a result on the other 3 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index56
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| MMLU Pro | 87.3% | 86.4% | DeepSeek V4 Pro 0423 |
| LegalBench (Vals AI) | 80.3% | 85.7% | GPT-5.1 |
| CaseLaw (Vals AI) | 59.4% | 73.4% | GPT-5.1 |
| CorpFin (Vals AI) | 61.4% | 63.8% | GPT-5.1 |
| TaxEval (Vals AI) | 72.1% | 74.9% | GPT-5.1 |
| MortgageTax (Vals AI) | Not measured | 61.4% | Only GPT-5.1 measured |
| MedQA (Vals AI) | Not measured | 96.4% | Only GPT-5.1 measured |
| MedCode (Vals AI) | 40.5% | 52.7% | GPT-5.1 |
| MMLU Pro (Artificial Analysis run) | Not measured | 87.0% | Only GPT-5.1 measured |
| SimpleQA Verified | 47.0% | 48.0% | GPT-5.1 |
| LMArena Text: science | 1470 | 1459 | DeepSeek V4 Pro 0423 |
| LMArena Text: medicine and healthcare | 1466 | 1456 | DeepSeek V4 Pro 0423 |
| LMArena Text: legal and government | 1464 | 1459 | DeepSeek V4 Pro 0423 |
| LMArena Text: business and finance | 1443 | 1431 | DeepSeek V4 Pro 0423 |
| Neither model has a result on the other 3 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index62
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| Design Arena: websites | 1235 | 1187 | DeepSeek V4 Pro 0423 |
| Design Arena: UI components | 1221 | 1166 | DeepSeek V4 Pro 0423 |
| Design Arena: game dev | 1229 | 1179 | DeepSeek V4 Pro 0423 |
| Design Arena: data viz | 1203 | 1204 | GPT-5.1 |
| Design Arena: 3D | 1254 | 1061 | DeepSeek V4 Pro 0423 |
| Design Arena: all code categories | 1240 | 1171 | DeepSeek V4 Pro 0423 |
| Design Arena: SVG | 1139 | 1149 | GPT-5.1 |
| Design Arena: ASCII art | 1148 | 1118 | DeepSeek V4 Pro 0423 |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index165
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| Public Benefits Bench (Vals AI) | 62.9% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| SAGE (Vals AI) | Not measured | 43.2% | Only GPT-5.1 measured |
| MedScribe (Vals AI) | 75.1% | 88.1% | GPT-5.1 |
| IFBench | Not measured | 72.9% | Only GPT-5.1 measured |
| Long context reasoning (AA-LCR) | Not measured | 80.0% | Only GPT-5.1 measured |
| DTBench | 90.7% | 90.1% | DeepSeek V4 Pro 0423 |
| LMCA | 41.2% | 43.9% | GPT-5.1 |
| CL-bench | Not measured | 23.7% | Only GPT-5.1 measured |
| Surface Evolver Bench | 40.0% | Not measured | Only DeepSeek V4 Pro 0423 measured |
| CL-bench Life | 13.5% | 17.3% | GPT-5.1 |
| LMArena Text | 1451 | 1443 | DeepSeek V4 Pro 0423 |
| LMArena Text: instruction following | 1445 | 1443 | DeepSeek V4 Pro 0423 |
| LMArena Text: creative writing | 1446 | 1427 | DeepSeek V4 Pro 0423 |
| LMArena Vision | Not measured | 1250 | Only GPT-5.1 measured |
| LMArena Text: English | 1459 | 1451 | DeepSeek V4 Pro 0423 |
| LMArena Text: non-English | 1439 | 1431 | DeepSeek V4 Pro 0423 |
| LMArena Text: Chinese | 1481 | 1495 | GPT-5.1 |
| LMArena Text: German | 1458 | 1438 | DeepSeek V4 Pro 0423 |
| LMArena Text: Japanese | 1446 | 1453 | GPT-5.1 |
| LMArena Text: Korean | 1409 | 1401 | DeepSeek V4 Pro 0423 |
| LMArena Text: French | 1473 | 1450 | DeepSeek V4 Pro 0423 |
| LMArena Text: Spanish | 1457 | 1433 | DeepSeek V4 Pro 0423 |
| LMArena Text: Russian | 1453 | 1435 | DeepSeek V4 Pro 0423 |
| LMArena Text: Polish | 1461 | 1440 | DeepSeek V4 Pro 0423 |
| LMArena Text: multi-turn | 1468 | 1450 | DeepSeek V4 Pro 0423 |
| LMArena Text: longer queries | 1458 | 1448 | DeepSeek V4 Pro 0423 |
| LMArena Text: writing and language | 1447 | 1428 | DeepSeek V4 Pro 0423 |
| LMArena Text: entertainment and media | 1431 | 1415 | DeepSeek V4 Pro 0423 |
| LMArena Vision: English | Not measured | 1252 | Only GPT-5.1 measured |
| LMArena Vision: Chinese | Not measured | 1270 | Only GPT-5.1 measured |
| LMArena Vision: reading text | Not measured | 1255 | Only GPT-5.1 measured |
| LMArena Vision: humor | Not measured | 1263 | Only GPT-5.1 measured |
| LMArena Vision: captioning | Not measured | 1252 | Only GPT-5.1 measured |
| LMArena Vision: recognizing things | Not measured | 1232 | Only GPT-5.1 measured |
| LMArena Vision: creative writing | Not measured | 1266 | Only GPT-5.1 measured |
| LMArena Document | Not measured | 1403 | Only GPT-5.1 measured |
| Neither model has a result on the other 10 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | DeepSeek V4 Pro 0423 | GPT-5.1 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
DeepSeek V4 Pro 0423 costs less
Repository review
50K input + 3K output tokens
DeepSeek V4 Pro 0423 costs less
Agent loop
200K input + 10K output tokens
DeepSeek V4 Pro 0423 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | DeepSeek V4 Pro 0423 | OpenAI GPT-5.1 |
|---|---|---|
| Maker | DeepSeek | OpenAI |
| Weights | Open weights | No public weights listed |
| Released | April 24, 2026 | November 13, 2025 |
| Context window | 1.05M | 400K |
| Input $/Mtok | $0.9553 | $1.25 |
| Output $/Mtok | $1.9105 | $10.00 |
| Coding index | 35.3 | 53.3 |
| Independent measurements behind it | 1 | 2 |
| Cost per solved task | $0.57 | $1.02 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is DeepSeek V4 Pro 0423 better than GPT-5.1 for coding?
Pick GPT-5.1 for coding: it scores 53.3 on the Vibe Coding Index against 35.3 for DeepSeek V4 Pro 0423. Pick DeepSeek V4 Pro 0423 when output tokens cost 5.2x less, or when a solved task costs $0.57 against $1.02.
Which is cheaper, DeepSeek V4 Pro 0423 or GPT-5.1?
DeepSeek V4 Pro 0423 costs $0.9553 per million input tokens and $1.9105 per million output tokens; GPT-5.1 costs $1.25 and $10.00. Per solved SWE-bench Verified task, DeepSeek V4 Pro 0423 costs $0.57 and GPT-5.1 $1.02.
Which has the bigger context window, DeepSeek V4 Pro 0423 or GPT-5.1?
DeepSeek V4 Pro 0423, with 1.05M tokens against 400K.
How many benchmarks is the DeepSeek V4 Pro 0423 vs GPT-5.1 comparison based on?
63 benchmarks have a result for both models as of October 21, 2026. DeepSeek V4 Pro 0423's coding index rests on 1 independent benchmark and GPT-5.1's on 2. Where only one model has a result, the page says so instead of naming a winner.