Microsoft model product
Phi 4
Phi 4 is an open large language model from Microsoft for reasoning, coding, and knowledge tasks.
Updated Sep 8, 2026. Default version: Phi 4
LLMBoard Score7.5Phi 4
Coverage40%13 benchmark families
Context window16KTokens
Official input price$0.125Azure API
Phi 4 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Phi 4 Benchmark Results
Benchmark scores for Phi 4.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkPhiBench | Score56.20% | Rank03 | Participants3 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkHumanEval+ | Score82.80% | Rank05 | Participants10 | Percentile55.56% | EvidenceC | Evaluated |
| BenchmarkArena Hard | Score75.40% | Rank09 | Participants26 | Percentile68.00% | EvidenceC | Evaluated |
| BenchmarkDROP | Score75.50% | Rank19 | Participants30 | Percentile37.93% | EvidenceC | Evaluated |
| BenchmarkMATH | Score80.40% | Rank19 | Participants71 | Percentile74.29% | EvidenceC | Evaluated |
| BenchmarkMGSM | Score80.60% | Rank19 | Participants31 | Percentile40.00% | EvidenceC | Evaluated |
| BenchmarkLiveBench | Score47.60% | Rank36 | Participants38 | Percentile5.41% | EvidenceC | Evaluated |
| BenchmarkHumanEval | Score82.60% | Rank41 | Participants66 | Percentile38.46% | EvidenceC | Evaluated |
| BenchmarkMMLU | Score84.80% | Rank45 | Participants101 | Percentile56.00% | EvidenceC | Evaluated |
| BenchmarkSimpleQA | Score3.00% | Rank45 | Participants47 | Percentile4.35% | EvidenceC | Evaluated |
| BenchmarkIFEval | Score63.00% | Rank66 | Participants68 | Percentile2.99% | EvidenceC | Evaluated |
| BenchmarkMMLU-Pro | Score70.40% | Rank89 | Participants138 | Percentile35.77% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score56.10% | Rank177 | Participants247 | Percentile28.46% | EvidenceC | Evaluated |
| BenchmarkLM Arena Text | Score1,216.47 rating | Rank196 | Participants210 | Percentile6.70% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Style Control | Score1,255.98 rating | Rank199 | Participants210 | Percentile5.26% | EvidenceA | Evaluated |
Phi 4 Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext | Categoryjapanese | Rank153 | Rating / score1,158.40 | Votes500 | ObservationsN/A | Result date |
| Arenatext style control | Categoryjapanese | Rank158 | Rating / score1,194.11 | Votes500 | ObservationsN/A | Result date |
| Arenatext | Categorykorean | Rank160 | Rating / score1,150.74 | Votes319 | ObservationsN/A | Result date |
| Arenatext | Categorygerman | Rank162 | Rating / score1,222.17 | Votes523 | ObservationsN/A | Result date |
| Arenatext style control | Categorykorean | Rank164 | Rating / score1,195.86 | Votes319 | ObservationsN/A | Result date |
| Arenatext | Categoryspanish | Rank165 | Rating / score1,233.53 | Votes127 | ObservationsN/A | Result date |
| Arenatext style control | Categorygerman | Rank167 | Rating / score1,255.68 | Votes523 | ObservationsN/A | Result date |
| Arenatext style control | Categoryspanish | Rank167 | Rating / score1,269.28 | Votes127 | ObservationsN/A | Result date |
| Arenatext | Categoryfrench | Rank168 | Rating / score1,223.96 | Votes205 | ObservationsN/A | Result date |
| Arenatext style control | Categoryfrench | Rank170 | Rating / score1,265.74 | Votes205 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry mathematical | Rank176 | Rating / score1,253.84 | Votes2,350 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry mathematical | Rank177 | Rating / score1,278.30 | Votes2,350 | ObservationsN/A | Result date |
| Arenatext | Categorymath | Rank178 | Rating / score1,245.70 | Votes2,764 | ObservationsN/A | Result date |
| Arenatext style control | Categorymath | Rank179 | Rating / score1,264.71 | Votes2,764 | ObservationsN/A | Result date |
| Arenatext | Categorychinese | Rank186 | Rating / score1,211.42 | Votes1,503 | ObservationsN/A | Result date |
| Arenatext | Categoryexpert | Rank187 | Rating / score1,202.80 | Votes1,124 | ObservationsN/A | Result date |
| Arenatext style control | Categoryexpert | Rank187 | Rating / score1,267.87 | Votes1,124 | ObservationsN/A | Result date |
| Arenatext style control | Categorychinese | Rank188 | Rating / score1,267.21 | Votes1,503 | ObservationsN/A | Result date |
| Arenatext | Categorycoding | Rank189 | Rating / score1,231.63 | Votes3,305 | ObservationsN/A | Result date |
| Arenatext style control | Categoryhard prompts english | Rank189 | Rating / score1,289.12 | Votes3,804 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank190 | Rating / score1,219.45 | Votes5,747 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts english | Rank190 | Rating / score1,230.01 | Votes3,804 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry legal and government | Rank190 | Rating / score1,241.04 | Votes1,341 | ObservationsN/A | Result date |
| Arenatext style control | Categorycoding | Rank190 | Rating / score1,306.18 | Votes3,305 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry legal and government | Rank191 | Rating / score1,298.77 | Votes1,341 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry software and it services | Rank192 | Rating / score1,228.13 | Votes5,580 | ObservationsN/A | Result date |
| Arenatext | Categoryinstruction following | Rank192 | Rating / score1,201.32 | Votes9,162 | ObservationsN/A | Result date |
| Arenatext style control | Categoryhard prompts | Rank192 | Rating / score1,277.49 | Votes5,747 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry medicine and healthcare | Rank193 | Rating / score1,201.02 | Votes1,037 | ObservationsN/A | Result date |
| Arenatext | Categorymulti turn | Rank193 | Rating / score1,206.07 | Votes3,517 | ObservationsN/A | Result date |
Phi 4 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.125 input, $0.50 output per 1M
- Official provider
- Azure
- Lowest third-party
- From $0.07 input, $0.14 output per 1M via OpenRouter
- Tracked offerings
- 3
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderOpenRouter | Provider model IDmicrosoft/phi-4 | Regionglobal | Input / 1M$0.07 | Output / 1M$0.14 | Context16.4K | Updated |
| ProviderAzure Cognitive Services | Provider model IDphi-4 | Regionglobal | Input / 1M$0.125 | Output / 1M$0.50 | Context128K | Updated |
| ProviderAzure | Provider model IDphi-4 | Regionglobal | Input / 1M$0.125 | Output / 1M$0.50 | Context128K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Phi 4 Runtime Performance
Provider-specific output speed and catalog latency for Phi 4. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderDeepInfra | Output Speed33.00 tok/s | Catalog Latency0.20 s | Max Input16K | Max Output16K | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Phi 4 Specifications
Technical details for the model's default version.
- Version
- Phi 4
- Released
- Dec 12, 2024
- Knowledge cutoff
- Jun 1, 2024
- Parameters
- 14.7B
- Context window
- 16K
- Max output
- 16K
- Inputs
- text
- Outputs
- text
- Open weights
- Yes
- License
- MIT
Phi 4 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionPhi 4 | Released | LLMBoard7.48 | Parameters14.7B | Context16K | Max output16K | Open weightsYes | LicenseMIT |
Models similar to Phi 4
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
MI
Phi 4 multimodal
Microsoft
MI
Phi 3.5 MoE
Microsoft
MI
Phi 4 Mini Reasoning
Microsoft
MI
Phi 3.5 mini
Microsoft
MI
Phi 3.5 vision
Microsoft
MI
Phi 4 Mini
Microsoft
What is Phi 4?
Key information about Phi 4 and its available data.
Phi 4 is an open model from Microsoft built for advanced reasoning, coding, and knowledge tasks. It was developed using synthetic data, filtered web data, academic texts, and supervised fine-tuning.
Data as of 2026-09-08.
FAQ
Common questions about Phi 4.
When was Phi 4 released?
Phi 4's default version was released on Dec 12, 2024.
How much does Phi 4 cost?
Phi 4's official API price is $0.125 per million input tokens and $0.50 per million output tokens via Azure. The lowest tracked third-party offer starts at $0.07 input and $0.14 output via OpenRouter.
Who created Phi 4?
Phi 4 was created by Microsoft.
What is the context window for Phi 4?
The default version has a 16K token context window.
Is Phi 4 open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer Phi 4?
3 provider offerings are linked to the default version.