Model comparison
DeepSeek-V2.5 (Sep 2024) vs Gemini 1.5 Pro (May 2024)
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 32.1 on the Noometry Index.
Last verified . 21 shared benchmarks.
DeepSeek
GoogleSummary
- They share 21 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 3 categories and Gemini 1.5 Pro (May 2024) in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek-V2.5 (Sep 2024) leads 25.6 to 12.3.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 37.6 | 32.1 |
| Released | 2024-09-06 | 2024-02-15 |
| Weights | Open | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 45 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 1.5 Pro (May 2024) leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Gemini 1.5 Pro (May 2024): 34.2 (#241)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| BigCodeBench Instruct | 48.6% | 43.8% |
| LMArena Coding | 1309 | 1294 |
| BigCodeBench Complete | 53.2% | 57.5% |
| HumanEval+ | 83.5% | 79.3% |
| MBPP+ | 74.1% | 74.6% |
| Aider Polyglot | 17.8% | — |
| WeirdML | — | 22.2% |
| CadEval | — | 34% |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Gemini 1.5 Pro (May 2024): 17.9 (#145)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| TheAgentCompany | — | 3.4% |
| Cybench | — | 7.5% |
| BALROG | — | 21% |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Gemini 1.5 Pro (May 2024): 12.3 (#338)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1296 |
| ARC-AGI-2 | — | 0.8% |
| SimpleBench | — | 27.1% |
| DTBench | — | 59% |
| BIG-Bench Hard | — | 89.2% |
| Epoch Capabilities Index | — | 131.73 |
| ForecastBench | — | 58.4 |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Gemini 1.5 Pro (May 2024): 25.8 (#266)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Math | 1288 | 1315 |
| OTIS Mock AIME 2024-2025 | — | 23.1% |
| Omni-MATH | — | 36.4% |
| MATH Level 5 | — | 70.4% |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Gemini 1.5 Pro (May 2024): 29.4 (#239)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Expert | 1266 | 1279 |
| GPQA Diamond | — | 57.2% |
| Humanity's Last Exam | — | 4.6% |
| MMLU-Pro | — | 73.7% |
| Confabulations | — | 13.5% |
| GPQA (HELM) | — | 53.4% |
| MMLU | — | 86.9% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Gemini 1.5 Pro (May 2024): 36.8 (#77)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Vision | — | 1161 |
| Video-MME | — | 75% |
Multilingual Gemini 1.5 Pro (May 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Gemini 1.5 Pro (May 2024): 45.3 (#174)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Non-English | 1273 | 1312 |
| LMArena Chinese | 1318 | 1331 |
| LMArena French | 1289 | 1302 |
| LMArena German | 1258 | 1286 |
| LMArena Japanese | 1228 | 1292 |
| LMArena Korean | 1209 | 1298 |
| LMArena Russian | 1289 | 1320 |
| LMArena Spanish | 1248 | 1311 |
Instruction Following Gemini 1.5 Pro (May 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Gemini 1.5 Pro (May 2024): 68.6 (#185)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Instruction Following | 1280 | 1297 |
| IFEval | — | 83.7% |
Long Context Too close to call
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Gemini 1.5 Pro (May 2024): 39.8 (#169)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Longer Query | 1301 | 1308 |
Writing & Preference Gemini 1.5 Pro (May 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Gemini 1.5 Pro (May 2024): 52.4 (#172)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 1.5 Pro (May 2024) |
|---|---|---|
| LMArena Text | 1294 | 1319 |
| LMArena Creative Writing | 1285 | 1333 |
| LMArena Multi-Turn | 1297 | 1296 |
| WildBench | — | 81.3% |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Gemini 1.5 Pro (May 2024)?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 32.1 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Gemini 1.5 Pro (May 2024) better for coding?
Gemini 1.5 Pro (May 2024) scores higher on coding benchmarks: 34.2 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Gemini 1.5 Pro (May 2024) share?
21 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Gemini 1.5 Pro (May 2024) has 45.