Phi-3-mini-4k-instruct
Microsoft- Arena score
- 1080
- Among open models
- #138of 148
- Among all on Arena
- #383
- Votes
- 12,297
Measured on Phi-3 Mini 4K Instruct (June 2024), the served version, not a local build. LMArena, CC BY 4.0, published Nov 11, 2026.
Score by category
CategoryScoreOpen rankArena rankVotes
- Math1152#116#3341.6K
- Applied math1147#120#3381.4K
- English1129#133#3726.7K
- Hard prompts, English1114#130#3632K
- Coding1093#132#3691.8K
- Software & IT1088#135#3743.1K
- Hard prompts1087#131#3663.2K
- Spanish1085#105#295410
- French1076#99#290307
- Legal & government1071#134#3691.1K
- Science1068#137#3792K
- Instructions1058#135#3784.4K
- Writing & literature1057#137#3893.4K
- Expert1051#129#353486
- Multi-turn1049#136#3811.8K
- Creative writing1044#137#3882K
- Long prompts1044#138#3811.7K
- German1044#102#305699
- Medicine & health1042#131#366653
- Chinese1034#130#3671.2K
- Business & finance1032#137#3792.5K
- Media & sports1031#137#3882K
- Russian1022#133#3712.8K
- Non-English1021#137#3839.8K
- Japanese955#89#281354
- Korean906#95#285860
- Excluding ties874#138#3888.2K
| Model | Params | Score |
|---|---|---|
| 3.8B | 1050 | |
| 3.8B | 1080 |
Built on it
Ways to run it
- GGUF · llama.cpp, Ollama, LM Studiolegraphista/Phi-3-mini-4k-instruct-update2024_07_03-IMat-GGUF137 builds
- MLX · Apple silicon: mlx-lm, LM Studiorogerscuall/phi3-mini-yoda-adapter-mac5 builds
- AWQ · vLLM, SGLangsolidrust/Phi-3-mini-4k-instruct-AWQ1 build
- GPTQ · vLLM, SGLangarnabsanyal/LazyWeights-Phi-3-mini-4k-instruct1 build
Most downloaded
- Phi-3-mini-4k-instruct-update2024_07_03-IMat-GGUFlegraphista · Quantization7.5K downloads
- Phi-3.1-mini-4k-instruct-GGUFbartowski · Quantization4.7K downloads
- Phi-3-mini-4k-instruct-llamafilemozilla-ai · Fine-tune3.9K downloads
- Phi3-ITA-mini-4K-instructe-palmisano · Fine-tune3.3K downloads
- Phi-3-mini-4k-instruct-GGUFsecond-state · Quantization2.9K downloads
BenchClub runs
Not runnable yet
Runs open once a Phi-3-mini-4k-instruct build is registered.
More from Microsoft
See more
Phi-3.5-mini-instruct324.4K3.8B

Phi-3-mini-128k-instruct241.1K3.8B

phi-4205.7K14.7B

phi-2177.4K2.8B
Other 3–9B models
See more
NVIDIA-Nemotron-3-Nano-4B-BF16NVIDIA1.9M4.0B
granite-4.1-8bIBM Granite79.4K8.8B
Molmo2-8BAi247.6K8.7B
Spark-X2.6-4BXHToken32.9K4.1B