Model comparison
DeepSeek LLM 67B vs phi-3-medium 14B
phi-3-medium 14B is the stronger model overall, scoring 29.7 to 24.9 on the Noometry Index.
Last verified . 3 shared benchmarks.
DeepSeek
MicrosoftSummary
- They share 3 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 0 categories and phi-3-medium 14B in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where phi-3-medium 14B leads 27.3 to 8.7.
- The biggest single-benchmark swing is MATH Level 5: 6.4% for DeepSeek LLM 67B and 17.6% for phi-3-medium 14B.
Side by side
| DeepSeek LLM 67B | phi-3-medium 14B | |
|---|---|---|
| Provider | DeepSeek | Microsoft |
| Noometry Index | 24.9 | 29.7 |
| Released | 2023-11-29 | 2024-04-23 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 15 | 13 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding phi-3-medium 14B leads
DeepSeek LLM 67B: 31.9 (#278), phi-3-medium 14B: 36.8 (#201)
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1096 | — |
| BigCodeBench Complete | — | 48.7% |
Reasoning Not comparable
DeepSeek LLM 67B: 16.5 (#304), phi-3-medium 14B: —
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| Epoch Capabilities Index | 110.5 | 121.23 |
| Chess Puzzles | 0% | — |
| LMArena Hard Prompts | 1070 | — |
| Adversarial NLI | — | 55.8% |
| BIG-Bench Hard | — | 81.4% |
| HellaSwag | — | 82.4% |
| WinoGrande | — | 81.5% |
Math phi-3-medium 14B leads
DeepSeek LLM 67B: 8.7 (#324), phi-3-medium 14B: 27.3 (#250)
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| MATH Level 5 | 6.4% | 17.6% |
| OTIS Mock AIME 2024-2025 | 0.8% | — |
| LMArena Math | 1108 | — |
Knowledge phi-3-medium 14B leads
DeepSeek LLM 67B: 7.0 (#313), phi-3-medium 14B: 9.1 (#306)
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| GPQA Diamond | 24.6% | 27.6% |
| ARC (AI2) Challenge | — | 91.6% |
| MMLU | — | 78% |
| OpenBookQA | — | 87.4% |
| TriviaQA | — | 73.9% |
Multilingual Not comparable
DeepSeek LLM 67B: 29.4 (#267), phi-3-medium 14B: —
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| LMArena Non-English | 1073 | — |
| LMArena Chinese | 1132 | — |
Instruction Following Not comparable
DeepSeek LLM 67B: 55.4 (#277), phi-3-medium 14B: —
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| LMArena Instruction Following | 1079 | — |
Long Context Not comparable
DeepSeek LLM 67B: 33.1 (#265), phi-3-medium 14B: —
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| LMArena Longer Query | 1092 | — |
Writing & Preference Not comparable
DeepSeek LLM 67B: 31.6 (#282), phi-3-medium 14B: —
| Benchmark | DeepSeek LLM 67B | phi-3-medium 14B |
|---|---|---|
| LMArena Text | 1105 | — |
| LMArena Creative Writing | 1067 | — |
| LMArena Multi-Turn | 1082 | — |
Frequently asked questions
Is DeepSeek LLM 67B better than phi-3-medium 14B?
phi-3-medium 14B is the stronger model overall, scoring 29.7 to 24.9 on the Noometry Index.
Is DeepSeek LLM 67B or phi-3-medium 14B better for coding?
phi-3-medium 14B scores higher on coding benchmarks: 36.8 versus 31.9 in the Noometry coding category.
How many benchmarks do DeepSeek LLM 67B and phi-3-medium 14B share?
3 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and phi-3-medium 14B has 13.