3models tracked · ranked by intelligence, speed & price
Microsoft's smartest model is Phi 4 (10.4 on the Intelligence Index), its fastest is Phi 4 Mini Instruct at 276 tokens/sec, and its cheapest is Phi 4 at $0.065 per million input tokens. All 3 Microsoft models are compared below by intelligence, speed, latency, context and price.
| Model | Intel | Speed | Latency | In $/M | Context |
|---|---|---|---|---|---|
| phi-4 JSON | 10.4 | 58 | 263ms | $0.065 | 16K |
| phi-4-mini-instruct JSON | 8.4 | 276 | 150ms | $0.080 | 131K |
| wizardlm-2-8x22b JSON | — | 10 | 1.0s | $0.620 | 66K |