Compare / head-to-head
Phi-4vsPhi-4 Reasoning
Phi-4 Reasoning leads 3 of 3 shared benchmarks. Phi-4 Reasoning has the larger context window (33K tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 3
Phi-4 Reasoning leads
Cheaper per token
—
list price, input + output
Larger context
Phi-4 Reasoning
33K tokens
Provider uptime (30d)
—
Microsoft
Phi-4
Microsoft · released 2024-12-12
text
- Context
- 16K
- Max out
- 16K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 11 · 7 core
Phi-4 Reasoning
Microsoft · released 2025-04-30
text
- Context
- 33K
- Max out
- 33K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 4 · 4 core
Quality
Benchmark matrix
3 shared · 7 only Phi-4 · 1 only Phi-4 Reasoning| Benchmark | Phi-4 | Phi-4 Reasoning | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 3 | ||||
| GPQA Diamond | 56.1% ↗ | 65.8% ↗ | -9.7 pt | Phi-4 Reasoning |
| MMLU-Pro | 70.4% ↗ | 74.3% ↗ | -3.9 pt | Phi-4 Reasoning |
| HumanEval+ | 82.8% ↗ | 92.9% ↗ | -10.1 pt | Phi-4 Reasoning |
| Only Phi-4 reports · 7 | ||||
| MMLU | 84.8% ↗ | not reported | — | — |
| HumanEval | 82.6% ↗ | not reported | — | — |
| LMArena Elo | 1216.6 ↗ | not reported | — | — |
| ArenaHard | 75.4% ↗ | not reported | — | — |
| IFEval | 63% ↗ | not reported | — | — |
| MATH | 80.4% ↗ | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 13.8% ↗epoch | not reported | — | — |
| Only Phi-4 Reasoning reports · 1 | ||||
| AIME 2025 | not reported | 62.9% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Phi-4 minus Phi-4 Reasoning in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Phi-4 | Phi-4 Reasoning | Edge |
|---|---|---|---|
| Context window | 16K tokens | 33K tokens | Phi-4 Reasoning |
| Max output | 16K tokens | 33K tokens | Phi-4 Reasoning |
| Input price / 1M | — | — | — |
| Output price / 1M | — | — | — |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | — | — | — |
| Modalities | text | text | Tie |
| Released | 2024-12-12 | 2025-04-30 | — |
| Cited benchmark scores | 11 | 4 | — |
Reliability
Microsoft status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Phi-4 vs …
Models sharing the most benchmarksMore matchups
Phi-4 Reasoning vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Phi-4 and Phi-4 Reasoning on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.