Model comparison
DeepSeek Coder 33B vs phi-3-medium 14B
phi-3-medium 14B has enough public results to be ranked (#306); DeepSeek Coder 33B does not yet, so treat this comparison as directional.
Last verified . 6 shared benchmarks.
DeepSeek
MicrosoftSummary
- They share 6 benchmarks with published results for both. DeepSeek Coder 33B scores higher in 1 category and phi-3-medium 14B in 0 categories; one gap is clear of the uncertainty.
Side by side
| DeepSeek Coder 33B | phi-3-medium 14B | |
|---|---|---|
| Provider | DeepSeek | Microsoft |
| Noometry Index | 38.9 | 29.7 |
| Released | 2023-11-02 | 2024-04-23 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 9 | 13 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek Coder 33B leads
DeepSeek Coder 33B: 38.0 (#184), phi-3-medium 14B: 36.8 (#201)
| Benchmark | DeepSeek Coder 33B | phi-3-medium 14B |
|---|---|---|
| BigCodeBench Instruct | 42% | 37.6% |
| BigCodeBench Complete | 51.1% | 48.7% |
| HumanEval+ | 75% | — |
| MBPP+ | 70.1% | — |
Reasoning Not comparable
DeepSeek Coder 33B: —, phi-3-medium 14B: —
| Benchmark | DeepSeek Coder 33B | phi-3-medium 14B |
|---|---|---|
| Epoch Capabilities Index | 96.32 | 121.23 |
| WinoGrande | 62% | 81.5% |
| Adversarial NLI | — | 55.8% |
| BIG-Bench Hard | — | 81.4% |
| HellaSwag | — | 82.4% |
Math Not comparable
DeepSeek Coder 33B: —, phi-3-medium 14B: 27.3 (#250)
| Benchmark | DeepSeek Coder 33B | phi-3-medium 14B |
|---|---|---|
| MATH Level 5 | — | 17.6% |
| GSM8K | 35.4% | — |
Knowledge Not comparable
DeepSeek Coder 33B: —, phi-3-medium 14B: 9.1 (#306)
| Benchmark | DeepSeek Coder 33B | phi-3-medium 14B |
|---|---|---|
| ARC (AI2) Challenge | 42.2% | 91.6% |
| MMLU | 39.4% | 78% |
| GPQA Diamond | — | 27.6% |
| OpenBookQA | — | 87.4% |
| TriviaQA | — | 73.9% |
Frequently asked questions
Is DeepSeek Coder 33B better than phi-3-medium 14B?
phi-3-medium 14B has enough public results to be ranked (#306); DeepSeek Coder 33B does not yet, so treat this comparison as directional.
Is DeepSeek Coder 33B or phi-3-medium 14B better for coding?
DeepSeek Coder 33B scores higher on coding benchmarks: 38.0 versus 36.8 in the Noometry coding category.
How many benchmarks do DeepSeek Coder 33B and phi-3-medium 14B share?
6 benchmarks have published results for both models. DeepSeek Coder 33B has 9 scored results on Noometry and phi-3-medium 14B has 13.