Model comparison
DeepSeek-V3.1-Terminus vs Phi 3 Mini 4k Instruct
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 43.1 to 27.9 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
MicrosoftSummary
- They share 10 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 7 categories and Phi 3 Mini 4k Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.1-Terminus leads 61.0 to 27.6.
Side by side
| DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct | |
|---|---|---|
| Provider | DeepSeek | Microsoft |
| Noometry Index | 43.1 | 27.9 |
| Released | 2025-09-22 | 2024-04-23 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 147K | — |
| Input $ / M tokens | $0.27 | — |
| Output $ / M tokens | $1 | — |
| Results tracked | 16 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 42.0 (#113), Phi 3 Mini 4k Instruct: 26.6 (#323)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Coding | 1426 | 1093 |
| SciCode | 40.6% | — |
| LiveBench Coding | — | 15.5% |
| ALE-Bench | 745.17 | — |
| HumanEval+ | — | 59.1% |
| MBPP+ | — | 54.2% |
Reasoning DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 26.4 (#133), Phi 3 Mini 4k Instruct: 14.1 (#328)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Hard Prompts | 1426 | 1072 |
| Kagi LLM Benchmark | 57.4% | — |
| CritPt | 1.7% | — |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | — | 26.8% |
| DTBench | 81.3% | — |
| LiveBench Data Analysis | — | 34.7% |
| LMCA | 28.6% | — |
| Adversarial NLI | — | 52.8% |
| BIG-Bench Hard | — | 71.7% |
| HellaSwag | — | 76.7% |
| LiveBench | — | 22.4% |
| WinoGrande | — | 70.8% |
Math DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 38.5 (#137), Phi 3 Mini 4k Instruct: 26.6 (#257)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Math | 1402 | 1111 |
| LiveBench Math | — | 15.7% |
Knowledge Not comparable
DeepSeek-V3.1-Terminus: —, Phi 3 Mini 4k Instruct: 28.5 (#246)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Expert | — | 1045 |
| ARC (AI2) Challenge | — | 84.9% |
| MMLU | — | 68.8% |
| OpenBookQA | — | 88% |
| TriviaQA | — | 64% |
Multilingual DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 52.1 (#92), Phi 3 Mini 4k Instruct: 26.3 (#280)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Non-English | 1407 | 1021 |
| LMArena Russian | 1436 | 1022 |
| LMArena Chinese | — | 1021 |
| LMArena French | — | 1076 |
| LMArena German | — | 1044 |
| LMArena Japanese | — | 935 |
| LMArena Korean | — | 905 |
| LMArena Spanish | — | 1085 |
Instruction Following DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 74.0 (#106), Phi 3 Mini 4k Instruct: 47.7 (#303)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Instruction Following | 1404 | 1053 |
| LiveBench Instruction Following | — | 39.1% |
Long Context DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 43.4 (#97), Phi 3 Mini 4k Instruct: 31.7 (#276)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Longer Query | 1421 | 1044 |
Writing & Preference DeepSeek-V3.1-Terminus leads
DeepSeek-V3.1-Terminus: 61.0 (#92), Phi 3 Mini 4k Instruct: 27.6 (#300)
| Benchmark | DeepSeek-V3.1-Terminus | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Text | 1419 | 1073 |
| LMArena Creative Writing | 1403 | 1037 |
| LMArena Multi-Turn | 1411 | 1018 |
| LiveBench Language | — | 9.2% |
Frequently asked questions
Is DeepSeek-V3.1-Terminus better than Phi 3 Mini 4k Instruct?
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 43.1 to 27.9 on the Noometry Index.
Is DeepSeek-V3.1-Terminus or Phi 3 Mini 4k Instruct better for coding?
DeepSeek-V3.1-Terminus scores higher on coding benchmarks: 42.0 versus 26.6 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.1-Terminus and Phi 3 Mini 4k Instruct share?
10 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Phi 3 Mini 4k Instruct has 35.