Quality
Highest overall scores
- 1
Microsoft: Phi 40
Microsoft model benchmarks.
Compare Microsoft models across answer quality, coding, complex tasks, speed, price, and context size.
Response speed
Fastest output in this sample
- 1Microsoft: Phi 465 t/s
Input price
Lowest price per 1M tokens
- 1Microsoft: Phi 4$0.07/1M
Highlights
Quality
Overall capability · Higher is better
Speed
Output tokens per second · Higher is better
Input price
USD per 1M tokens · Lower is better
Quality benchmarks
Compare Microsoft model capability.
Performance and price
Compare practical tradeoffs.
All Microsoft models.
1 model in the current catalogue.
| Model | Links | |||||
|---|---|---|---|---|---|---|
| Microsoft: Phi 4 | 0 | 65 t/s | 16.4K | $0.07 | $0.14 |
Microsoft technical benchmarks.
Compare named evaluation records available for Microsoft models. Missing results remain blank rather than being estimated.
| Benchmark | Area | Phi 4 | Source |
|---|---|---|---|
| GPQA | GPQA Diamond | 57.5 | OpenRouter model benchmarks |
| Humanity's Last Exam | HLE | 4.1 | OpenRouter model benchmarks |
| IFBench | IFBench | 23.5 | OpenRouter model benchmarks |
| τ²-Bench Telecom | τ²-Bench Telecom | 0 | OpenRouter model benchmarks |
| AA-LCR | AA-LCR | 0 | OpenRouter model benchmarks |
| CritPt | CritPt | 0 | OpenRouter model benchmarks |
| SciCode | SciCode | 26 | OpenRouter model benchmarks |
| Terminal-Bench Hard | Terminal-Bench Hard | 3.8 | OpenRouter model benchmarks |
| AA-Omniscience Accuracy | AA-Omniscience Accuracy | 13.2 | OpenRouter model benchmarks |
| AA-Omniscience Non-Hallucination Rate | AA-Omniscience Non-Hallucination Rate | 19.5 | OpenRouter model benchmarks |
Microsoft image models.
| Model | Quality | Evidence | Price /1k images | Links |
|---|---|---|---|---|
| MAI-Image-2.5 | 1270 Elo | 10,073 samples | $48.1 | View details |
| MAI-Image-2.5 Pro | — | — | — | View details |