Compare / head-to-head
Phi-4-mini-flash-reasoningvs
Qwen3 4B Instruct 2507
Qwen3 4B Instruct 2507 leads 2 of 2 shared benchmarks. Qwen3 4B Instruct 2507 has the larger context window (262K tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 2
Qwen3 4B Instruct 2507 leads
Cheaper per token
—
list price, input + output
Larger context
Qwen3 4B Instruct 2507
262K tokens
Providers
2 providers
Microsoft · Qwen
Phi-4-mini-flash-reasoning
Microsoft · released 2025-07-09
text
- Context
- 64K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 4 · 3 core
Qwen3 4B Instruct 2507
Qwen · released 2025-08-06
text
- Context
- 262K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 18 · 7 core
Quality
Benchmark matrix
2 shared · 2 only Phi-4-mini-flash-reasoning · 16 only Qwen3 4B Instruct 2507| Benchmark | Phi-4-mini-flash-reasoning | Qwen3 4B Instruct 2507 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 2 | ||||
| GPQA Diamond | 45.08% ↗ | 45.8% ↗epoch run | -0.7 pt | Qwen3 4B Instruct 2507 |
| AIME25 | 33.59% ↗ | 47.4% ↗ | -13.8 pt | Qwen3 4B Instruct 2507 |
| Only Phi-4-mini-flash-reasoning reports · 2 | ||||
| AIME24 | 52.29% ↗ | not reported | — | — |
| Math500 | 92.45% ↗ | not reported | — | — |
| Only Qwen3 4B Instruct 2507 reports · 16 | ||||
| MMLU-Pro | not reported | 69.6% ↗ | — | — |
| Aider-Polyglot | not reported | 12.9% ↗ | — | — |
| Arena-Hard v2 | not reported | 43.4% ↗ | — | — |
| BFCL-v3 | not reported | 61.9% ↗ | — | — |
| Chess Puzzles (Epoch AI run) | not reported | 4% ↗epoch run | — | — |
| GPQA | not reported | 62% ↗ | — | — |
| HMMT25 | not reported | 31% ↗ | — | — |
| IFEval | not reported | 83.4% ↗ | — | — |
| LiveBench 20241125 | not reported | 63% ↗ | — | — |
| LiveCodeBench v6 (25.02-25.05) | not reported | 35.1% ↗ | — | — |
| MMLU-Redux | not reported | 84.2% ↗ | — | — |
| MultiPL-E | not reported | 76.8% ↗ | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | not reported | 52.2% ↗epoch run | — | — |
| SuperGPQA | not reported | 42.8% ↗ | — | — |
| TAU2-Retail | not reported | 40.4% ↗ | — | — |
| ZebraLogic | not reported | 80.2% ↗ | — | — |
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-mini-flash-reasoning minus Qwen3 4B Instruct 2507 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-mini-flash-reasoning | Qwen3 4B Instruct 2507 | Edge |
|---|---|---|---|
| Context window | 64K tokens | 262K tokens | Qwen3 4B Instruct 2507 |
| Max output | — | — | — |
| 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 | 2025-07-09 | 2025-08-06 | — |
| Cited benchmark scores | 4 | 18 | — |
Reliability
Provider status
Live from /statusMore matchups
Phi-4-mini-flash-reasoning vs …
Models sharing the most benchmarksMore matchups
Qwen3 4B Instruct 2507 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Phi-4-mini-flash-reasoning and Qwen3 4B Instruct 2507 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.