- Size
- 3.8B parameterscounted from the safetensors header on the Hub
- Licence
- custom, read it, MIT LicenseOur reading, check the licence: A custom licence we have not read for this model: other makers use the same name for different texts. Read the model's own licence file before you use it.Licence text
- On the Hub since
- 19 Feb 2025the date the repo was created
- Context length
- 128k tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)Phi3ForCausalLM
- On Hugging Face
- microsoft/Phi-4-mini-instruct305k downloads in the last 30 days, 892 likes
Scores
- default setting0.0%via Epoch AI
Epoch's name: Phi-4-mini-instruct. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Independent scores from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Each row links its own source. What the badges mean.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 1.0 GB, worked out from the model's config (microsoft/Phi-4-mini-instruct), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
unsloth/Phi-4-mini-instruct-GGUF (converted by unsloth), 119k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| Q2_K | 1.6 GBfile size | 3.1 GB1.6 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q2_KOllama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q2_KDownload hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct-Q2_K.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| Q2_K_L | 1.6 GBfile size | 3.1 GB1.6 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q2_K_LOllama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q2_K_LDownload hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct-Q2_K_L.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| Q3_K_M | 2.0 GBfile size | 3.5 GB2.0 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q3_K_MOllama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q3_K_MDownload hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct-Q3_K_M.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| Q4_K_M | 2.3 GBfile size | 3.8 GB2.3 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q4_K_MOllama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q4_K_MDownload hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct-Q4_K_M.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| Q5_K_M | 2.7 GBfile size | 4.2 GB2.7 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q5_K_MOllama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q5_K_MDownload hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct-Q5_K_M.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| Q6_K | 2.9 GBfile size | 4.4 GB2.9 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q6_KOllama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q6_KDownload hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct-Q6_K.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| Q8_0 | 3.8 GBfile size | 5.3 GB3.8 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:Q8_0Ollama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:Q8_0Download hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct.Q8_0.gguf" --local-dir Phi-4-mini-instruct-GGUF |
| BF16 | 7.2 GBfile size | 8.7 GB7.2 + 1.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Phi-4-mini-instruct-GGUF:BF16Ollama ollama run hf.co/unsloth/Phi-4-mini-instruct-GGUF:BF16Download hf download unsloth/Phi-4-mini-instruct-GGUF --include "Phi-4-mini-instruct.BF16.gguf" --local-dir Phi-4-mini-instruct-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX 4bitmlx-community/Phi-4-mini-instruct-4bit | 2.0 GBfile size | 3.5 GB2.0 + 1.0 + 0.50 | Download page |
| MLX 8bitmlx-community/Phi-4-mini-instruct-8bit | 3.8 GBfile size | 5.3 GB3.8 + 1.0 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 7.1 GBfile size | 8.6 GB7.1 + 1.0 + 0.50 |
hf download microsoft/Phi-4-mini-instruct --local-dir Phi-4-mini-instructOther conversions (15; sizes not read)
- unsloth/Phi-4-mini-instruct-unsloth-bnb-4bit bitsandbytes, 4bit, 4.1k downloads
- unsloth/Phi-4-mini-instruct-bnb-4bit bitsandbytes, 4bit, 2.4k downloads
- OpenVINO/Phi-4-mini-instruct-int4-ov Other, int4, 2.2k downloads
- RedHatAI/Phi-4-mini-instruct-FP8-dynamic FP8, 2.2k downloads
- pytorch/Phi-4-mini-instruct-INT8-INT4 Other, int8, 855 downloads
- iqbalamo93/Phi-4-mini-instruct-GPTQ-8bit GPTQ, 8bit, 413 downloads
- onnx-community/Phi-4-mini-instruct-ONNX Other, 394 downloads
- jamalshipit/Phi-4-mini-instruct-int4-ov Other, int4, 252 downloads
- pytorch/Phi-4-mini-instruct-FP8 FP8, 156 downloads
- pytorch/Phi-4-mini-instruct-AWQ-INT4 AWQ, int4, 139 downloads
- drawais/Phi-4-mini-instruct-AWQ-INT4 AWQ, int4, 96 downloads
- Natsukiz/Phi-4-mini-instruct-bnb-4bit bitsandbytes, 4bit, 40 downloads
- simaai/Phi-4-mini-instruct-Autoround-Safetensors GPTQ, 39 downloads
- aleada/Phi-4-mini-instruct-W4A16 GPTQ, w4a16, 34 downloads
- ErtasAI/Phi-4-mini-instruct-bnb-4bit bitsandbytes, 4bit, 34 downloads
The Hub tags 100 or more quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
No provider on Hugging Face Inference Providers serves this model (as of 19 Nov 2026, 13:58 UTC). Other hosts may; we only list what the router reports.
Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Fine-tunes of this model
Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.
- litert-community/Phi-4-mini-instruct24kUnreviewed
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.