Model comparison
DeepSeek-V3.1 vs Gemini 3.8 Flash
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 42.8 on the Noometry Index. DeepSeek-V3.1 costs 3.5× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
DeepSeek
GoogleSummary
- They share 22 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 0 categories and Gemini 3.8 Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 27.9.
- The biggest single-benchmark swing is WeirdML: 38.4% for DeepSeek-V3.1 and 84.8% for Gemini 3.8 Flash.
- DeepSeek-V3.1 is cheaper at $0.25 / $0.95 per million input/output tokens, against $0.75 / $3.75 for Gemini 3.8 Flash.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 164K.
- DeepSeek-V3.1 has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.1 | Gemini 3.8 Flash | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 42.8 | 61.8 |
| Released | 2025-08-21 | 2026-09-02 |
| Weights | Open | Proprietary |
| Context window | 164K | 1.05M |
| Max output | 8K | 66K |
| Input $ / M tokens | $0.25 | $0.75 |
| Output $ / M tokens | $0.95 | $3.75 |
| Results tracked | 27 | 50 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.8 Flash leads
DeepSeek-V3.1: 40.3 (#144), Gemini 3.8 Flash: 59.2 (#15)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| WeirdML | 38.4% | 84.8% |
| LMArena Coding | 1417 | 1510 |
| DeepSWE | — | 73.8% |
| FrontierCode | — | 41.2% |
| CursorBench | — | 39.6% |
| LMArena WebDev | — | 1584 |
| FrontierSWE | — | 19.6% |
| SciCode | — | 56.6% |
| ALE-Bench | — | 1,270 |
Agentic & Tool Use Not comparable
DeepSeek-V3.1: —, Gemini 3.8 Flash: 41.8 (#21)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| APEX-Agents | — | 64.3% |
| Remote Labor Index | — | 5.8% |
| GDP.pdf | — | 23.4% |
| Vending-Bench 2 | — | 5,094 |
Reasoning Gemini 3.8 Flash leads
DeepSeek-V3.1: 27.9 (#110), Gemini 3.8 Flash: 76.9 (#5)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Hard Prompts | 1417 | 1508 |
| DTBench | 82.7% | 95.7% |
| LMCA | 24.3% | 52.9% |
| Epoch Capabilities Index | 139.92 | 156.71 |
| ARC-AGI-2 | — | 89.2% |
| SimpleBench | 40% | — |
| Kagi LLM Benchmark | 53.2% | — |
| NYT Connections (extended) | — | 97.4% |
| ARC-AGI-1 | — | 98.5% |
| CritPt | — | 18.3% |
| Chess Puzzles | — | 61% |
| Mystery Game Puzzles | — | 47% |
| Surface Evolver Bench | — | 76.9% |
| ForecastBench | 58 | — |
Math Gemini 3.8 Flash leads
DeepSeek-V3.1: 38.9 (#122), Gemini 3.8 Flash: 65.3 (#28)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Math | 1420 | 1528 |
| FrontierMath (Tiers 1-3) | — | 68.4% |
| FrontierMath Tier 4 | — | 22% |
| OTIS Mock AIME 2024-2025 | — | 98.9% |
| ProofBench | — | 48% |
Knowledge Gemini 3.8 Flash leads
DeepSeek-V3.1: 43.7 (#90), Gemini 3.8 Flash: 74.8 (#2)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Expert | 1405 | 1524 |
| GPQA Diamond | — | 95.4% |
| Humanity's Last Exam | — | 44.5% |
| SimpleQA Verified | — | 69.7% |
| Vectara Hallucination Rate | 5.5% | — |
Multimodal Not comparable
DeepSeek-V3.1: —, Gemini 3.8 Flash: 40.7 (#45)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Vision | — | 1314 |
| Blueprint-Bench 2 | — | 38.6% |
| Furniture Assembly | — | 31.7% |
Multilingual Gemini 3.8 Flash leads
DeepSeek-V3.1: 51.6 (#106), Gemini 3.8 Flash: 58.0 (#5)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Non-English | 1400 | 1491 |
| LMArena Chinese | 1469 | 1554 |
| LMArena French | 1447 | 1498 |
| LMArena German | 1411 | 1493 |
| LMArena Japanese | 1378 | 1502 |
| LMArena Korean | 1337 | 1459 |
| LMArena Russian | 1405 | 1515 |
| LMArena Spanish | 1431 | 1485 |
Instruction Following Gemini 3.8 Flash leads
DeepSeek-V3.1: 73.9 (#110), Gemini 3.8 Flash: 78.0 (#13)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Instruction Following | 1400 | 1490 |
Long Context Gemini 3.8 Flash leads
DeepSeek-V3.1: 36.3 (#232), Gemini 3.8 Flash: 46.3 (#24)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Longer Query | 1422 | 1508 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference Gemini 3.8 Flash leads
DeepSeek-V3.1: 60.3 (#98), Gemini 3.8 Flash: 72.2 (#15)
| Benchmark | DeepSeek-V3.1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Text | 1420 | 1499 |
| LMArena Creative Writing | 1401 | 1492 |
| EQ-Bench Creative Writing | 1436 | 1748 |
| LMArena Multi-Turn | 1408 | 1501 |
Frequently asked questions
Is DeepSeek-V3.1 better than Gemini 3.8 Flash?
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 42.8 on the Noometry Index. DeepSeek-V3.1 costs 3.5× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3.1 or Gemini 3.8 Flash?
DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Gemini 3.8 Flash lists at $0.75 and $3.75.
Is DeepSeek-V3.1 or Gemini 3.8 Flash better for coding?
Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 40.3 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.8 Flash does, with 1.05M tokens against 164K.
How many benchmarks do DeepSeek-V3.1 and Gemini 3.8 Flash share?
22 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Gemini 3.8 Flash has 50.