Model comparison
DeepSeek-V2.5 (Sep 2024) vs Gemini 3.6 Flash
Gemini 3.6 Flash is the stronger model overall, scoring 54.1 to 37.6 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
GoogleSummary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 0 categories and Gemini 3.6 Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.6 Flash leads 58.8 to 25.6.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 37.6 | 54.1 |
| Released | 2024-09-06 | 2026-07-21 |
| Weights | Open | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 66K |
| Input $ / M tokens | — | $0.75 |
| Output $ / M tokens | — | $3.75 |
| Results tracked | 22 | 46 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Gemini 3.6 Flash: 50.0 (#48)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Coding | 1309 | 1491 |
| DeepSWE | — | 46.7% |
| FrontierCode | — | 34.4% |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1538 |
| SciCode | — | 52.7% |
| WeirdML | — | 56.1% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| ALE-Bench | — | 715.52 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Gemini 3.6 Flash: 32.3 (#65)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| APEX-Agents | — | 46.9% |
| GDP.pdf | — | 14% |
Reasoning Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Gemini 3.6 Flash: 58.8 (#22)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1485 |
| ARC-AGI-2 | — | 60.4% |
| NYT Connections (extended) | — | 89% |
| ARC-AGI-1 | — | 91.2% |
| CritPt | — | 10.6% |
| Chess Puzzles | — | 43% |
| Mystery Game Puzzles | — | 30% |
| DTBench | — | 95.5% |
| LMCA | — | 44.9% |
| Epoch Capabilities Index | — | 154.25 |
Math Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Gemini 3.6 Flash: 57.3 (#40)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Math | 1288 | 1505 |
| FrontierMath (Tiers 1-3) | — | 58.9% |
| FrontierMath Tier 4 | — | 22% |
| MathArena Final-Answer Competitions | — | 70.8% |
| OTIS Mock AIME 2024-2025 | — | 94.2% |
| ProofBench | — | 36% |
Knowledge Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Gemini 3.6 Flash: 67.8 (#8)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Expert | 1266 | 1488 |
| GPQA Diamond | — | 94.1% |
| SimpleQA Verified | — | 66.2% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Gemini 3.6 Flash: 38.5 (#64)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Vision | — | 1298 |
| Blueprint-Bench 2 | — | 31.2% |
| Furniture Assembly | — | 23.3% |
| LMArena Document | — | 1456 |
Multilingual Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Gemini 3.6 Flash: 56.5 (#19)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Non-English | 1273 | 1469 |
| LMArena Chinese | 1318 | 1531 |
| LMArena French | 1289 | 1504 |
| LMArena German | 1258 | 1478 |
| LMArena Japanese | 1228 | 1476 |
| LMArena Korean | 1209 | 1431 |
| LMArena Russian | 1289 | 1487 |
| LMArena Spanish | 1248 | 1475 |
Instruction Following Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Gemini 3.6 Flash: 77.0 (#33)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Instruction Following | 1280 | 1466 |
Long Context Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Gemini 3.6 Flash: 45.1 (#50)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Longer Query | 1301 | 1474 |
Writing & Preference Gemini 3.6 Flash leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Gemini 3.6 Flash: 68.2 (#27)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.6 Flash |
|---|---|---|
| LMArena Text | 1294 | 1479 |
| LMArena Creative Writing | 1285 | 1465 |
| LMArena Multi-Turn | 1297 | 1481 |
| EQ-Bench Creative Writing | — | 1604 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Gemini 3.6 Flash?
Gemini 3.6 Flash is the stronger model overall, scoring 54.1 to 37.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Gemini 3.6 Flash better for coding?
Gemini 3.6 Flash scores higher on coding benchmarks: 50.0 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Gemini 3.6 Flash share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Gemini 3.6 Flash has 46.