DeepSeek V3.2 vs Gemini 3.1 Flash Lite Preview
DeepSeek V3.2 leads Gemini 3.1 Flash Lite Preview on coding, 30.0 to 22.7 on the Vibe Coding Index, and DeepSeek V3.2 costs 1.7x less per output token ($0.88 vs $1.50 per million). Updated October 22, 2026; 51 benchmarks have a result for both.
DeepSeek V3.2 or Gemini 3.1 Flash Lite Preview: which to pick
Pick DeepSeek V3.2 for coding: it scores 30.0 on the Vibe Coding Index against 22.7 for Gemini 3.1 Flash Lite Preview.
Pick Gemini 3.1 Flash Lite Preview for a lower cost per solved task ($0.18 against $0.99) and a larger context window (1.05M against 164K tokens).
- Shared results
- 51
- DeepSeek V3.2 only
- 24
- Gemini 3.1 Flash Lite Preview only
- 28
- Like-for-like categories
- 7 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
DeepSeek V3.2
DeepSeek V3.2 scores 30.0 against 22.7.
3 benchmarks behind each score
Intelligence
AA Intelligence Index
DeepSeek V3.2
DeepSeek V3.2 leads, 21.5 against 15.6.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Gemini 3.1 Flash Lite Preview has a AA Agentic Index result, 1.6.
Directional only
Reasoning
GPQA Diamond
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview leads, 81.1% against 80.3%.
Same board, both measured
Math
AIME
DeepSeek V3.2
DeepSeek V3.2 leads, 84.6% against 83.3%.
Same board, both measured
Knowledge
MMLU Pro
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview leads, 86.2% against 84.9%.
Same board, both measured
Web design
Design Arena: websites
DeepSeek V3.2
DeepSeek V3.2 leads, 1175 against 1082.
Same board, both measured
Long documents
Listed context window
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview takes 1.05M tokens against 164K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
DeepSeek V3.2
DeepSeek V3.2 $0.0007, Gemini 3.1 Flash Lite Preview $0.0010 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
Gemini 3.1 Flash Lite Preview
DeepSeek V3.2 $0.99, Gemini 3.1 Flash Lite Preview $0.18.
Published attempt cost and pass rate
At a glance
- Coding index
- DeepSeek V3.2 30.0 from 3 benchmarks, Gemini 3.1 Flash Lite Preview 22.7 from 3 benchmarks.
- Widest category
- Web design, 10.4 points apart on average over 8 shared benchmarks.
- Widest single gap
- LMArena WebDev: React: DeepSeek V3.2 1350 against Gemini 3.1 Flash Lite Preview 1245, 17.4 points apart on a 0 to 100 scale.
- Output price
- DeepSeek V3.2 $0.88 against Gemini 3.1 Flash Lite Preview $1.50 per million output tokens, 1.7x apart.
- Context window
- DeepSeek V3.2 takes 164K tokens, Gemini 3.1 Flash Lite Preview 1.05M.
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 | Gemini 3.1 Flash Lite Preview | Ties |
|---|---|---|---|---|
| Coding | 9 | 0 | ||
| Intelligence | 1 | 0 | ||
| Reasoning | 3 | 1 | ||
| Math | 3 | 0 | ||
| Knowledge | 5 | 4 | ||
| Web design | 6 | 2 | ||
| General | 9 | 8 | ||
| All shared boards | 36 | 15 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index90
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| SWE-bench Verified | 70.0% | Not measured | Only DeepSeek V3.2 measured |
| SWE-bench Multilingual | 59.0% | Not measured | Only DeepSeek V3.2 measured |
| LMArena WebDev | 1362 | 1257 | DeepSeek V3.2 |
| AA Coding Index | 44.2 | 34.7 | DeepSeek V3.2 |
| LiveCodeBench (Vals AI) | 80.7% | 80.1% | DeepSeek V3.2 |
| SWE-bench Verified (Vals AI) | 67.6% | 62.8% | DeepSeek V3.2 |
| Vibe Code Bench (Vals AI) | 5.1% | 0.0% | DeepSeek V3.2 |
| Code migration (Vals AI) | Not measured | 4.6% | Only Gemini 3.1 Flash Lite Preview measured |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Gemini 3.1 Flash Lite Preview measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 34.1% | Only Gemini 3.1 Flash Lite Preview measured |
| LiveCodeBench (Artificial Analysis run) | 86.2% | Not measured | Only DeepSeek V3.2 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 35.6% | Not measured | Only DeepSeek V3.2 measured |
| Terminal-Bench | 39.6% | Not measured | Only DeepSeek V3.2 measured |
| ALE-Bench | Not measured | 797.7 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Text: coding | 1453 | 1400 | DeepSeek V3.2 |
| LMArena Text: software and IT | 1449 | 1413 | DeepSeek V3.2 |
| LMArena WebDev: React | 1350 | 1245 | DeepSeek V3.2 |
| LMArena WebDev: HTML | 1371 | 1276 | DeepSeek V3.2 |
| LMArena WebDev: from an image | Not measured | 1336 | Only Gemini 3.1 Flash Lite Preview measured |
| SciCode | Not measured | 4.6% | Only Gemini 3.1 Flash Lite Preview measured |
| Neither model has a result on the other 19 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 21.5 | 15.6 | DeepSeek V3.2 |
| Epoch Capabilities Index | 146.1 | Not measured | Only DeepSeek V3.2 measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 1.6 | Only Gemini 3.1 Flash Lite Preview measured |
| Finance Agent (Vals AI) | Not measured | 30.0% | Only Gemini 3.1 Flash Lite Preview measured |
| Tax Agent Bench (Vals AI) | Not measured | 3.1% | Only Gemini 3.1 Flash Lite Preview measured |
| Legal Research Bench (Vals AI) | Not measured | 3.4% | Only Gemini 3.1 Flash Lite Preview measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 8.6% | Only Gemini 3.1 Flash Lite Preview measured |
| τ²-Bench Telecom | 90.6% | Not measured | Only DeepSeek V3.2 measured |
| APEX-Agents | 21.3% | Not measured | Only DeepSeek V3.2 measured |
| Vending-Bench 2 | $1,034 | Not measured | Only DeepSeek V3.2 measured |
| Neither model has a result on the other 33 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index31
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| GPQA Diamond | 80.3% | 81.1% | Gemini 3.1 Flash Lite Preview |
| MMMU Pro (Vals AI) | Not measured | 82.5% | Only Gemini 3.1 Flash Lite Preview measured |
| GPQA Diamond (Artificial Analysis run) | 84.0% | Not measured | Only DeepSeek V3.2 measured |
| Humanity's Last Exam (Artificial Analysis run) | 24.6% | Not measured | Only DeepSeek V3.2 measured |
| ARC-AGI | 57.0% | Not measured | Only DeepSeek V3.2 measured |
| ARC-AGI-2 | 4.0% | Not measured | Only DeepSeek V3.2 measured |
| LMArena Text: hard prompts | 1434 | 1407 | DeepSeek V3.2 |
| LMArena Text: expert | 1436 | 1398 | DeepSeek V3.2 |
| LMArena Text: hard prompts in English | 1446 | 1403 | DeepSeek V3.2 |
| LMArena Vision: diagrams | Not measured | 1242 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: homework | Not measured | 1271 | Only Gemini 3.1 Flash Lite Preview measured |
| Neither model has a result on the other 11 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| AIME | 84.6% | 83.3% | DeepSeek V3.2 |
| MGSM (Vals AI) | 86.0% | Not measured | Only DeepSeek V3.2 measured |
| AA Math Index | 92.0 | Not measured | Only DeepSeek V3.2 measured |
| AIME 2025 (Artificial Analysis run) | 92.0% | Not measured | Only DeepSeek V3.2 measured |
| LMArena Text: math | 1435 | 1428 | DeepSeek V3.2 |
| LMArena Text: mathematical work | 1440 | 1417 | DeepSeek V3.2 |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index54
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| MMLU Pro | 84.9% | 86.2% | Gemini 3.1 Flash Lite Preview |
| LegalBench (Vals AI) | 76.1% | 83.8% | Gemini 3.1 Flash Lite Preview |
| CaseLaw (Vals AI) | 55.4% | 55.0% | DeepSeek V3.2 |
| CorpFin (Vals AI) | 51.0% | 59.4% | Gemini 3.1 Flash Lite Preview |
| TaxEval (Vals AI) | 68.2% | 71.8% | Gemini 3.1 Flash Lite Preview |
| MortgageTax (Vals AI) | Not measured | 68.0% | Only Gemini 3.1 Flash Lite Preview measured |
| MedQA (Vals AI) | 93.9% | Not measured | Only DeepSeek V3.2 measured |
| MedCode (Vals AI) | Not measured | 47.6% | Only Gemini 3.1 Flash Lite Preview measured |
| MMLU Pro (Artificial Analysis run) | 86.2% | Not measured | Only DeepSeek V3.2 measured |
| LMArena Text: science | 1446 | 1425 | DeepSeek V3.2 |
| LMArena Text: medicine and healthcare | 1440 | 1428 | DeepSeek V3.2 |
| LMArena Text: legal and government | 1425 | 1418 | DeepSeek V3.2 |
| LMArena Text: business and finance | 1410 | 1389 | DeepSeek V3.2 |
| FACTS Benchmark Suite | Not measured | 51.4% | Only Gemini 3.1 Flash Lite Preview measured |
| Neither model has a result on the other 4 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index62
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1175 | 1082 | DeepSeek V3.2 |
| Design Arena: UI components | 1146 | 1069 | DeepSeek V3.2 |
| Design Arena: game dev | 1131 | 1029 | DeepSeek V3.2 |
| Design Arena: data viz | 1158 | 1045 | DeepSeek V3.2 |
| Design Arena: 3D | 1132 | 1047 | DeepSeek V3.2 |
| Design Arena: all code categories | 1163 | 1072 | DeepSeek V3.2 |
| Design Arena: SVG | 1032 | 1052 | Gemini 3.1 Flash Lite Preview |
| Design Arena: ASCII art | 1083 | 1167 | Gemini 3.1 Flash Lite Preview |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index98
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | Leader |
|---|---|---|---|
| SAGE (Vals AI) | Not measured | 49.5% | Only Gemini 3.1 Flash Lite Preview measured |
| MedScribe (Vals AI) | Not measured | 63.9% | Only Gemini 3.1 Flash Lite Preview measured |
| Vals Multimodal Index | Not measured | 41.4% | Only Gemini 3.1 Flash Lite Preview measured |
| IFBench | 60.7% | Not measured | Only DeepSeek V3.2 measured |
| Long context reasoning (AA-LCR) | 73.3% | Not measured | Only DeepSeek V3.2 measured |
| DTBench | 85.6% | Not measured | Only DeepSeek V3.2 measured |
| LMCA | 28.8% | Not measured | Only DeepSeek V3.2 measured |
| CL-bench | 12.4% | Not measured | Only DeepSeek V3.2 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 | 1377 | DeepSeek V3.2 |
| LMArena Text: creative writing | 1400 | 1401 | Gemini 3.1 Flash Lite Preview |
| LMArena Vision | Not measured | 1240 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Text: English | 1439 | 1414 | DeepSeek V3.2 |
| LMArena Text: non-English | 1409 | 1410 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: Chinese | 1455 | 1460 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: German | 1413 | 1430 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: Japanese | 1374 | 1413 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: Korean | 1370 | 1392 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: French | 1429 | 1424 | DeepSeek V3.2 |
| LMArena Text: Spanish | 1424 | 1421 | DeepSeek V3.2 |
| LMArena Text: Russian | 1424 | 1420 | DeepSeek V3.2 |
| LMArena Text: Polish | 1419 | 1443 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: multi-turn | 1427 | 1417 | DeepSeek V3.2 |
| LMArena Text: longer queries | 1428 | 1394 | DeepSeek V3.2 |
| LMArena Text: writing and language | 1407 | 1408 | Gemini 3.1 Flash Lite Preview |
| LMArena Text: entertainment and media | 1395 | 1386 | DeepSeek V3.2 |
| LMArena Vision: English | Not measured | 1232 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: Chinese | Not measured | 1278 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: reading text | Not measured | 1246 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: humor | Not measured | 1266 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: captioning | Not measured | 1259 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: recognizing things | Not measured | 1308 | Only Gemini 3.1 Flash Lite Preview measured |
| LMArena Vision: creative writing | Not measured | 1248 | Only Gemini 3.1 Flash Lite Preview measured |
| Neither model has a result on the other 12 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | DeepSeek V3.2 | Gemini 3.1 Flash Lite Preview | 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
Gemini 3.1 Flash Lite Preview costs less
Agent loop
200K input + 10K output tokens
Gemini 3.1 Flash Lite Preview 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 | Google Gemini 3.1 Flash Lite Preview |
|---|---|---|
| Maker | DeepSeek | |
| Weights | Open weights | No public weights listed |
| Released | December 1, 2025 | March 3, 2026 |
| Context window | 164K | 1.05M |
| Input $/Mtok | $0.29 | $0.25 |
| Output $/Mtok | $0.88 | $1.50 |
| Coding index | 30.0 | 22.7 |
| Independent measurements behind it | 3 | 3 |
| Cost per solved task | $0.99 | $0.18 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is DeepSeek V3.2 better than Gemini 3.1 Flash Lite Preview for coding?
Pick DeepSeek V3.2 for coding: it scores 30.0 on the Vibe Coding Index against 22.7 for Gemini 3.1 Flash Lite Preview. Pick Gemini 3.1 Flash Lite Preview for a lower cost per solved task ($0.18 against $0.99) and a larger context window (1.05M against 164K tokens).
Which is cheaper, DeepSeek V3.2 or Gemini 3.1 Flash Lite Preview?
DeepSeek V3.2 costs $0.29 per million input tokens and $0.88 per million output tokens; Gemini 3.1 Flash Lite Preview costs $0.25 and $1.50. Per solved SWE-bench Verified task, DeepSeek V3.2 costs $0.99 and Gemini 3.1 Flash Lite Preview $0.18.
Which has the bigger context window, DeepSeek V3.2 or Gemini 3.1 Flash Lite Preview?
Gemini 3.1 Flash Lite Preview, with 1.05M tokens against 164K.
How many benchmarks is the DeepSeek V3.2 vs Gemini 3.1 Flash Lite Preview comparison based on?
51 benchmarks have a result for both models as of October 22, 2026. DeepSeek V3.2's coding index rests on 3 independent benchmarks and Gemini 3.1 Flash Lite Preview's on 3. Where only one model has a result, the page says so instead of naming a winner.