Phi 3.5 MoE Instruct is the stronger all-rounder.
Scores updated · 18 tests both models report · How we compare
Where each one wins
Tests won in each of the one area where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
Reasoning rests on a single test.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Phi 3.5 MoE Instruct pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+6.4points ahead
Where Phi 3.5 Mini Instruct pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 18 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningPhi 3.5 MoE Instruct
- GPQA DiamondPhi 3.5 MoE Instruct by 6.430.436.8+6.4
Other results17 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- MMMLUPhi 3.5 MoE Instruct by 14.555.469.9+14.5
- HellaSwagPhi 3.5 MoE Instruct by 14.469.483.8+14.4
- TruthfulQAPhi 3.5 MoE Instruct by 13.56477.5+13.5
- WinoGrandePhi 3.5 MoE Instruct by 12.868.581.3+12.8
- MBPPPhi 3.5 MoE Instruct by 11.269.680.8+11.2
- MGSMPhi 3.5 MoE Instruct by 10.847.958.7+10.8
- OpenBookQAPhi 3.5 MoE Instruct by 10.479.289.6+10.4
- BBHPhi 3.5 MoE Instruct by 10.16979.1+10.1
- HumanEvalPhi 3.5 MoE Instruct by 7.962.870.7+7.9
- PIQAPhi 3.5 MoE Instruct by 7.68188.6+7.6
- BoolQPhi 3.5 MoE Instruct by 6.67884.6+6.6
- ARC-ChallengePhi 3.5 MoE Instruct by 6.484.691+6.4
- Social IQAPhi 3.5 MoE Instruct by 3.374.778+3.3
- RULERPhi 3.5 MoE Instruct by 384.187.1+3
- GSM8KPhi 3.5 MoE Instruct by 2.586.288.7+2.5
- MMLU-ProPhi 3.5 Mini Instruct by 2.147.445.3+2.1
- Arena Hardtie3737.9tie
Questions people ask
Which is better, Phi 3.5 Mini Instruct or Phi 3.5 MoE Instruct?
Phi 3.5 MoE Instruct wins the one area where both have results: reasoning. Phi 3.5 Mini Instruct wins none.
How do you compare the two?
We use the 18 benchmark tests both models have published scores on. The verdict counts the 1 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 17 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.