DeepSeek V3.2 vs GPT-5.2
GPT-5.2 leads DeepSeek V3.2 on coding, 55.9 to 30.0 on the Vibe Coding Index, and DeepSeek V3.2 costs 15.9x less per output token ($0.88 vs $14.00 per million). Updated October 21, 2026; 71 benchmarks have a result for both.
DeepSeek V3.2 or GPT-5.2: which to pick
Pick GPT-5.2 for coding: it scores 55.9 on the Vibe Coding Index against 30.0 for DeepSeek V3.2.
Pick DeepSeek V3.2 for output tokens 15.9x cheaper ($0.88 vs $14.00 per million) and a lower cost per solved task ($0.99 against $1.91).
- Shared results
- 71
- DeepSeek V3.2 only
- 4
- GPT-5.2 only
- 43
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
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, six public coding benchmarks
GPT-5.2
GPT-5.2 scores 55.9 against 30.0.
3 benchmarks behind each score
Intelligence
AA Intelligence Index
GPT-5.2
GPT-5.2 leads, 30.4 against 21.5.
Same board, both measured
Reasoning
GPQA Diamond
GPT-5.2
GPT-5.2 leads, 91.7% against 80.3%.
Same board, both measured
Math
AIME
GPT-5.2
GPT-5.2 leads, 96.9% against 84.6%.
Same board, both measured
Knowledge
MMLU Pro
GPT-5.2
GPT-5.2 leads, 86.2% against 84.9%.
Same board, both measured
Web design
Design Arena: websites
GPT-5.2
GPT-5.2 leads, 1195 against 1175.
Same board, both measured
Long documents
Listed context window
GPT-5.2
GPT-5.2 takes 400K tokens against 164K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
DeepSeek V3.2
DeepSeek V3.2 $0.0007, GPT-5.2 $0.0088 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
DeepSeek V3.2
DeepSeek V3.2 $0.99, GPT-5.2 $1.91.
Published attempt cost and pass rate
At a glance
- Coding index
- DeepSeek V3.2 30.0 from 3 benchmarks, GPT-5.2 55.9 from 3 benchmarks.
- Widest category
- Reasoning, 14.2 points apart on average over 8 shared benchmarks.
- Widest single gap
- ARC-AGI-2: DeepSeek V3.2 4.0% against GPT-5.2 52.9%, 48.9 points apart on a 0 to 100 scale.
- Output price
- DeepSeek V3.2 $0.88 against GPT-5.2 $14.00 per million output tokens, 15.9x apart.
- Context window
- DeepSeek V3.2 takes 164K tokens, GPT-5.2 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 V3.2 | Bars | GPT-5.2 | Ties |
|---|---|---|---|---|
| Coding | 2 | 10 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 1 | 1 | ||
| Reasoning | 2 | 6 | ||
| Math | 0 | 6 | ||
| Knowledge | 2 | 9 | ||
| Web design | 1 | 7 | ||
| General | 12 | 10 | ||
| All shared boards | 20 | 51 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index210
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| SWE-bench Verified | 70.0% | 72.8% | GPT-5.2 |
| SWE-bench Multilingual | 59.0% | 66.7% | GPT-5.2 |
| LMArena WebDev | 1362 | 1417 | GPT-5.2 |
| AA Coding Index | 44.2 | Not measured | Only DeepSeek V3.2 measured |
| LiveCodeBench (Vals AI) | 80.7% | 85.4% | GPT-5.2 |
| SWE-bench Verified (Vals AI) | 67.6% | 75.8% | GPT-5.2 |
| Vibe Code Bench (Vals AI) | 5.1% | 53.5% | GPT-5.2 |
| LiveCodeBench (Artificial Analysis run) | 86.2% | 89.4% | GPT-5.2 |
| Terminal-Bench Hard (Artificial Analysis run) | 35.6% | 47.0% | GPT-5.2 |
| Terminal-Bench | 39.6% | 64.9% | GPT-5.2 |
| GSO-Bench | Not measured | 27.5% | Only GPT-5.2 measured |
| WeirdML | Not measured | 72.2% | Only GPT-5.2 measured |
| ALE-Bench | Not measured | 1293.5 | Only GPT-5.2 measured |
| LMArena Text: coding | 1453 | 1442 | DeepSeek V3.2 |
| LMArena Text: software and IT | 1449 | 1434 | DeepSeek V3.2 |
| LMArena WebDev: React | 1350 | Not measured | Only DeepSeek V3.2 measured |
| LMArena WebDev: HTML | 1371 | 1427 | GPT-5.2 |
| Neither model has a result on the other 22 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| AA Intelligence Index | 21.5 | 30.4 | GPT-5.2 |
| Epoch Capabilities Index | 146.1 | 153.7 | GPT-5.2 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| Design Arena: agents building web apps | Not measured | 1086 | Only GPT-5.2 measured |
| Design Arena: agents building full-stack apps | Not measured | 1043 | Only GPT-5.2 measured |
| Design Arena: agents building mobile apps | Not measured | 1093 | Only GPT-5.2 measured |
| Design Arena: agents building Android apps | Not measured | 1058 | Only GPT-5.2 measured |
| Design Arena: agents building Godot games | Not measured | 1142 | Only GPT-5.2 measured |
| τ²-Bench Telecom | 90.6% | 84.8% | DeepSeek V3.2 |
| APEX-Agents | 21.3% | Not measured | Only DeepSeek V3.2 measured |
| METR time horizons | Not measured | 75.3% | Only GPT-5.2 measured |
| Remote Labor Index | Not measured | 2.5% | Only GPT-5.2 measured |
| GDPval | Not measured | 49.7% | Only GPT-5.2 measured |
| DeepResearch Bench | Not measured | 41.1% | Only GPT-5.2 measured |
| Vending-Bench 2 | $1,034 | $3,591 | GPT-5.2 |
| Berkeley Function Calling Leaderboard (BFCL) | Not measured | 55.9% | Only GPT-5.2 measured |
| Neither model has a result on the other 28 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index26
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| GPQA Diamond | 80.3% | 91.7% | GPT-5.2 |
| MMMU Pro (Vals AI) | Not measured | 86.7% | Only GPT-5.2 measured |
| GPQA Diamond (Artificial Analysis run) | 84.0% | 90.3% | GPT-5.2 |
| Humanity's Last Exam (Artificial Analysis run) | 24.6% | 37.7% | GPT-5.2 |
| GPQA Diamond (Epoch AI run) | Not measured | 91.4% | Only GPT-5.2 measured |
| ARC-AGI | 57.0% | 86.2% | GPT-5.2 |
| ARC-AGI-2 | 4.0% | 52.9% | GPT-5.2 |
| SimpleBench | Not measured | 45.8% | Only GPT-5.2 measured |
| Chess puzzles | Not measured | 48.0% | Only GPT-5.2 measured |
| Mystery game puzzles | Not measured | 23.0% | Only GPT-5.2 measured |
| EBR-bench | Not measured | 23.0% | Only GPT-5.2 measured |
| VPCT | Not measured | 84.0% | Only GPT-5.2 measured |
| LMArena Text: hard prompts | 1434 | 1428 | DeepSeek V3.2 |
| LMArena Text: expert | 1436 | 1445 | GPT-5.2 |
| LMArena Text: hard prompts in English | 1446 | 1431 | DeepSeek V3.2 |
| LMArena Vision: diagrams | Not measured | 1265 | Only GPT-5.2 measured |
| LMArena Vision: homework | Not measured | 1276 | Only GPT-5.2 measured |
| Kagi LLM Benchmark | Not measured | 73.3% | Only GPT-5.2 measured |
| Neither model has a result on the other 4 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| AIME | 84.6% | 96.9% | GPT-5.2 |
| MGSM (Vals AI) | 86.0% | 94.0% | GPT-5.2 |
| AA Math Index | 92.0 | 99.0 | GPT-5.2 |
| AIME 2025 (Artificial Analysis run) | 92.0% | 99.0% | GPT-5.2 |
| OTIS Mock AIME | Not measured | 96.1% | Only GPT-5.2 measured |
| FrontierMath, tiers 1 to 3 | Not measured | 67.4% | Only GPT-5.2 measured |
| FrontierMath, tier 4 | Not measured | 31.7% | Only GPT-5.2 measured |
| ProofBench | Not measured | 15.0% | Only GPT-5.2 measured |
| LMArena Text: math | 1435 | 1440 | GPT-5.2 |
| LMArena Text: mathematical work | 1440 | 1442 | GPT-5.2 |
| Neither model has a result on the other 3 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index29
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| MMLU Pro | 84.9% | 86.2% | GPT-5.2 |
| LegalBench (Vals AI) | 76.1% | 82.8% | GPT-5.2 |
| CaseLaw (Vals AI) | 55.4% | 66.0% | GPT-5.2 |
| CorpFin (Vals AI) | 51.0% | 65.9% | GPT-5.2 |
| TaxEval (Vals AI) | 68.2% | 75.8% | GPT-5.2 |
| MortgageTax (Vals AI) | Not measured | 67.1% | Only GPT-5.2 measured |
| MedQA (Vals AI) | 93.9% | 94.1% | GPT-5.2 |
| MedCode (Vals AI) | Not measured | 49.8% | Only GPT-5.2 measured |
| MMLU Pro (Artificial Analysis run) | 86.2% | 87.4% | GPT-5.2 |
| SimpleQA Verified | Not measured | 37.1% | Only GPT-5.2 measured |
| LMArena Text: science | 1446 | 1425 | DeepSeek V3.2 |
| LMArena Text: medicine and healthcare | 1440 | 1425 | DeepSeek V3.2 |
| LMArena Text: legal and government | 1425 | 1427 | GPT-5.2 |
| LMArena Text: business and finance | 1410 | 1414 | GPT-5.2 |
| FACTS Benchmark Suite | Not measured | 48.5% | Only GPT-5.2 measured |
| Neither model has a result on the other 3 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index17
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| Design Arena: websites | 1175 | 1195 | GPT-5.2 |
| Design Arena: UI components | 1149 | 1190 | GPT-5.2 |
| Design Arena: game dev | 1132 | 1195 | GPT-5.2 |
| Design Arena: data viz | 1158 | 1202 | GPT-5.2 |
| Design Arena: 3D | 1133 | 1079 | DeepSeek V3.2 |
| Design Arena: all code categories | 1164 | 1170 | GPT-5.2 |
| Design Arena: SVG | 1032 | 1139 | GPT-5.2 |
| Design Arena: ASCII art | 1083 | 1152 | GPT-5.2 |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index1210
| Benchmark | DeepSeek V3.2 | GPT-5.2 | Leader |
|---|---|---|---|
| SAGE (Vals AI) | Not measured | 49.3% | Only GPT-5.2 measured |
| MedScribe (Vals AI) | Not measured | 84.4% | Only GPT-5.2 measured |
| IFBench | 60.7% | 75.4% | GPT-5.2 |
| Long context reasoning (AA-LCR) | 73.3% | 82.7% | GPT-5.2 |
| DTBench | 85.6% | 90.9% | GPT-5.2 |
| LMCA | 28.8% | 43.9% | GPT-5.2 |
| CL-bench | 12.4% | 18.2% | GPT-5.2 |
| Furniture assembly | Not measured | 38.3% | Only GPT-5.2 measured |
| CL-bench Life | 7.4% | Not measured | Only DeepSeek V3.2 measured |
| LMArena Text | 1425 | 1416 | DeepSeek V3.2 |
| LMArena Text: instruction following | 1413 | 1409 | DeepSeek V3.2 |
| LMArena Text: creative writing | 1400 | 1376 | DeepSeek V3.2 |
| LMArena Vision | Not measured | 1246 | Only GPT-5.2 measured |
| LMArena Text: English | 1439 | 1422 | DeepSeek V3.2 |
| LMArena Text: non-English | 1409 | 1405 | DeepSeek V3.2 |
| LMArena Text: Chinese | 1455 | 1447 | DeepSeek V3.2 |
| LMArena Text: German | 1413 | 1444 | GPT-5.2 |
| LMArena Text: Japanese | 1374 | 1420 | GPT-5.2 |
| LMArena Text: Korean | 1370 | 1374 | GPT-5.2 |
| LMArena Text: French | 1429 | 1454 | GPT-5.2 |
| LMArena Text: Spanish | 1424 | 1413 | DeepSeek V3.2 |
| LMArena Text: Russian | 1424 | 1415 | DeepSeek V3.2 |
| LMArena Text: Polish | 1419 | 1426 | GPT-5.2 |
| LMArena Text: multi-turn | 1427 | 1419 | DeepSeek V3.2 |
| LMArena Text: longer queries | 1428 | 1413 | DeepSeek V3.2 |
| LMArena Text: writing and language | 1407 | 1392 | DeepSeek V3.2 |
| LMArena Text: entertainment and media | 1395 | 1378 | DeepSeek V3.2 |
| LMArena Vision: English | Not measured | 1243 | Only GPT-5.2 measured |
| LMArena Vision: Chinese | Not measured | 1305 | Only GPT-5.2 measured |
| LMArena Vision: reading text | Not measured | 1257 | Only GPT-5.2 measured |
| LMArena Vision: humor | Not measured | 1227 | Only GPT-5.2 measured |
| LMArena Vision: captioning | Not measured | 1267 | Only GPT-5.2 measured |
| LMArena Vision: recognizing things | Not measured | 1232 | Only GPT-5.2 measured |
| LMArena Vision: creative writing | Not measured | 1225 | Only GPT-5.2 measured |
| LMArena Document | Not measured | 1405 | Only GPT-5.2 measured |
| Neither model has a result on the other 11 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | DeepSeek V3.2 | GPT-5.2 | 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 V3.2 costs less
Repository review
50K input + 3K output tokens
DeepSeek V3.2 costs less
Agent loop
200K input + 10K output tokens
DeepSeek V3.2 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 V3.2 | OpenAI GPT-5.2 |
|---|---|---|
| Maker | DeepSeek | OpenAI |
| Weights | Open weights | No public weights listed |
| Released | December 1, 2025 | December 10, 2025 |
| Context window | 164K | 400K |
| Input $/Mtok | $0.29 | $1.75 |
| Output $/Mtok | $0.88 | $14.00 |
| Coding index | 30.0 | 55.9 |
| Independent measurements behind it | 3 | 3 |
| Cost per solved task | $0.99 | $1.91 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is DeepSeek V3.2 better than GPT-5.2 for coding?
Pick GPT-5.2 for coding: it scores 55.9 on the Vibe Coding Index against 30.0 for DeepSeek V3.2. Pick DeepSeek V3.2 for output tokens 15.9x cheaper ($0.88 vs $14.00 per million) and a lower cost per solved task ($0.99 against $1.91).
Which is cheaper, DeepSeek V3.2 or GPT-5.2?
DeepSeek V3.2 costs $0.29 per million input tokens and $0.88 per million output tokens; GPT-5.2 costs $1.75 and $14.00. Per solved SWE-bench Verified task, DeepSeek V3.2 costs $0.99 and GPT-5.2 $1.91.
Which has the bigger context window, DeepSeek V3.2 or GPT-5.2?
GPT-5.2, with 400K tokens against 164K.
How many benchmarks is the DeepSeek V3.2 vs GPT-5.2 comparison based on?
71 benchmarks have a result for both models as of October 21, 2026. DeepSeek V3.2's coding index rests on 3 independent benchmarks and GPT-5.2's on 3. Where only one model has a result, the page says so instead of naming a winner.