Model comparison
DeepSeek-V2.5 (Sep 2024) vs Gemini 3.5 Flash
Gemini 3.5 Flash is the stronger model overall, scoring 54.1 to 37.0 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 3.5 Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.5 Flash leads 62.7 to 24.5.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 37.0 | 54.1 |
| Released | 2024-09-06 | 2026-05-19 |
| Weights | Open | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 66K |
| Input $ / M tokens | — | $1.50 |
| Output $ / M tokens | — | $9 |
| Results tracked | 22 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), Gemini 3.5 Flash: 49.1 (#46)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Coding | 1310 | 1492 |
| SWE-bench Verified | — | 79.3% |
| DeepSWE | — | 37.4% |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1499 |
| SciCode | — | 53.1% |
| WeirdML | — | 62.6% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| ALE-Bench | — | 911.02 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Gemini 3.5 Flash: 24.8 (#107)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| APEX-Agents | — | 27.5% |
| GBAEval | — | 6.7% |
| GDP.pdf | — | 14% |
| Vending-Bench 2 | — | 5,396 |
Reasoning Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), Gemini 3.5 Flash: 62.7 (#16)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1488 |
| ARC-AGI-2 | — | 72.1% |
| SimpleBench | — | 76.7% |
| NYT Connections (extended) | — | 92.6% |
| ARC-AGI-1 | — | 92.5% |
| CritPt | — | 13.1% |
| Chess Puzzles | — | 50% |
| EnigmaEval | — | 25.4% |
| EBR-Bench | — | 4.8% |
| Mystery Game Puzzles | — | 32% |
| DTBench | — | 94.7% |
| LMCA | — | 47.1% |
| Surface Evolver Bench | — | 58.1% |
| Epoch Capabilities Index | — | 154.55 |
| ForecastBench | — | 59 |
Math Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), Gemini 3.5 Flash: 60.5 (#30)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Math | 1289 | 1502 |
| FrontierMath (Tiers 1-3) | — | 62.8% |
| FrontierMath Tier 4 | — | 26.8% |
| MathArena Final-Answer Competitions | — | 76.3% |
| OTIS Mock AIME 2024-2025 | — | 95.6% |
| ProofBench | — | 31% |
| FrontierMath (Feb 2025 set) | — | 39% |
| FrontierMath Tier 4 (v1) | — | 14.6% |
Knowledge Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), Gemini 3.5 Flash: 66.3 (#10)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Expert | 1266 | 1493 |
| GPQA Diamond | — | 92.8% |
| SimpleQA Verified | — | 66.2% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Gemini 3.5 Flash: 45.7 (#13)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Vision | — | 1311 |
| Blueprint-Bench 2 | — | 33.6% |
| LMArena Document | — | 1463 |
Multilingual Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), Gemini 3.5 Flash: 57.1 (#12)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Non-English | 1273 | 1475 |
| LMArena Chinese | 1317 | 1525 |
| LMArena French | 1288 | 1490 |
| LMArena German | 1258 | 1500 |
| LMArena Japanese | 1228 | 1481 |
| LMArena Korean | 1209 | 1453 |
| LMArena Russian | 1289 | 1491 |
| LMArena Spanish | 1248 | 1478 |
Instruction Following Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), Gemini 3.5 Flash: 77.2 (#29)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1280 | 1466 |
Long Context Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), Gemini 3.5 Flash: 45.3 (#39)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Longer Query | 1301 | 1481 |
Writing & Preference Gemini 3.5 Flash leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), Gemini 3.5 Flash: 65.5 (#46)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 3.5 Flash |
|---|---|---|
| LMArena Text | 1294 | 1481 |
| LMArena Creative Writing | 1285 | 1467 |
| LMArena Multi-Turn | 1297 | 1482 |
| EQ-Bench 4 | — | 1087 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Gemini 3.5 Flash?
Gemini 3.5 Flash is the stronger model overall, scoring 54.1 to 37.0 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Gemini 3.5 Flash better for coding?
Gemini 3.5 Flash scores higher on coding benchmarks: 49.1 versus 31.3 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Gemini 3.5 Flash share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Gemini 3.5 Flash has 54.