MicrosoftPhi-4-Mini
Phi-4-Mini is a Microsoft open-weights model, released 27 Feb 2025. Its best configuration ranks #761 of 785 on the BenchLeader Index at 35.8 ±8.1, in the lower half. The ± is the point: 227 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in reasoning (41) and lowest in coding (5). At $0.135 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 46 tokens per second puts it in the slowest quarter, with a first answer in 0.8 s. Last measured 27 Feb 2025.stale: no new result in six months
- Blended price
- $0.135/M
- $0.080 in · $0.300 out
- Output speed
- 46 tok/s
- measured by Artificial Analysis
- First answer
- 0.82 s
- first token 0.82 s
- Context
- 128k
- Full answer
- 12 s
- median, reasoning included
- Cost per run
- –
- Released
- 27 Feb 2025
- Overall index36
- Reasoning41
- Coding5
- Agents & tools34
- Knowledge34
- Instruction following26
- Long context31
- Composite37
Versions
Microsoft has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Phi-4-Minithis page | 27 Feb 2025 | 35.8 | #761 |
| Phi-3.5-mini | 16 Aug 2024 | – | – |
| Phi 3 Mini | 23 Apr 2024 | 36.8 | #743 |
How its index has moved
9 Sept 2026 to 21 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | not stated | Source |
|---|---|---|
| GPQA Diamond (AA)not in index | 33.1%#505 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 4.5%#503 | Artificial Analysis |
| CritPt | 0.0%#315 | Artificial Analysis |
Coding
| Benchmark | not stated | Source |
|---|---|---|
| SciCode | 10.8%#215 | SciCode |
| Terminal-Bench 2.1 (AA)not in index | 0.4%#228 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source |
|---|---|---|
| Terminal-Bench Hard | 0.0%#362 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 8.2%#389 | Artificial Analysis |
| GDPval-AA v2.1not in index | 0.0%#124 | Artificial Analysis |
Knowledge
| Benchmark | not stated | Source |
|---|---|---|
| AA-Omniscience | -61.1#471 | Artificial Analysis |
| AA-Omniscience: accuracynot in index | 9.5%#502 | Artificial Analysis |
| AA-Omniscience: non-hallucinationnot in index | 22.0%#267 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source |
|---|---|---|
| IFBench | 21.1%#411 | Artificial Analysis |
Long context
| Benchmark | not stated | Source |
|---|---|---|
| AA-LCR | 15.3%#453 | Artificial Analysis |
Composite
| Benchmark | not stated | Source |
|---|---|---|
| AA Intelligence Index v4.3.2 | 6.3#535 | Artificial Analysis |
Price history
The price above, per 1M tokens, every day it changed.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0001 | – | 7.4 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0011 | – | 14.0 s |
| Code edit | 6,000 / 1,500 | $0.0009 | – | 33.7 s |
| Agentic coding session | 60,000 / 4,000 | $0.0060 | – | 1.5 min |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 5.2 s |
More from Microsoft
See also
Questions people ask
- What are Phi-4-Mini's benchmark scores?
- As of 25 Oct 2026, Phi-4-Mini scores 35.8 ±8.1 on the BenchLeader Index, strongest in reasoning (41) and weakest in coding (5). That rests on 7 independent benchmark results, and every score on this page links to its publisher.
- How does Phi-4-Mini rank?
- Phi-4-Mini is #761 of 785 ranked configurations on 25 Oct 2026. 227 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Phi-4-Mini cost per 1M tokens?
- As of 25 Oct 2026, Phi-4-Mini costs $0.080 per 1M input tokens and $0.300 per 1M output tokens, $0.135 blended (three input tokens per output token). The list price last changed on 13 Oct 2026.
- When was Phi-4-Mini released?
- Microsoft released Phi-4-Mini on 27 Feb 2025.
Data as of 25 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Phi-4-Mini: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/phi-4-mini.