whatmodel.sh / Phi-1 (1.3B)

Phi-1 (1.3B)

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Phi-1 (1.3B) is an open-weights language model from Microsoft, released September 2023. Its published size is 1.42B parameters and a 2k-token context window. Licence: mit.

Catalogue identity notes

Phi-1 1.3B Python code model (Gunasekar et al. 2023), base model.

20.7±4.2#686 of 719 LLMsfrom 3 benchmark results across 2 sources

$ how it compares

The LLMs ranked either side of Phi-1 (1.3B), out of 828 with a score, then more from Microsoft.

#modelscorewhereRAMreleased
790BLOOM 7.1B BigScience21±1local~7.1GB2022-07
791OLMo 1B Allen Institute for AI21±2local~2.2GB2024-02
792Phi-1 (1.3B) Microsoft21±4local~2.4GB2023-09
793Qwen1.5 0.5B Chat Alibaba21±3local~1.8GB2024-01
794ChatGLM-6B Zhipu AI (THUDM)20±2local~16GB2023-03
more from Microsoft
252Phi-4-reasoning-plus 14B Microsoft59±1local~10GB2025-04
261Phi-4-reasoning 14B Microsoft58±1api $0.22/M +local~10GB2025-04
365Phi-4 14B Microsoft53±1api $0.087/M +local~10GB2024-12

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▃▃▆██▇▆▄▃▁ #448 of 458 * · 76% of its score · 4 tests

Beats Molmo 7B-D (0924), DeepSeek R1 Distill Qwen 1.5B, Qwen3 0.6B here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
CRUXEval-O21.7% ✎▃·▂▃▁█▁▂···▁ beats 0/22medium · avg 53%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 40%blunt
CRUXEval-I13.1% ✎▃··▁▅▂█▇▁··▁ beats 0/22medium · avg 54%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 37%typical
HumanEval lab50.6% ✎▁▁▁▁▁▃▅▅█▃▂▁ beats 81/190easy · avg 79%—not counted yet: no other model in the same run to compare it with
MBPP lab55.5% ✎▁▁▁▄▅█▄▅▃▁▂▁ beats 44/81easy · avg 70%—not counted yet: no other model in the same run to compare it with
other tests1 not in a task · 16% of its score
testresultvs the others in its rundifficultycountswhy
Open PT-LLM Leaderboard average (9 Portuguese tasks)27.5% ✎▃▁▁▁▁▁▁▁▂█▅▁ beats 35/276medium · avg 73%▇▇▇▇▇▇ 16%noisy

$ where to run it

llama.cpp · recommended

version

  1. Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
    macOS / Linux
    curl -LsSf https://llama.app/install.sh | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run Phi-1 (1.3B) · needs ~0.9 GB · from huggingface.co/mradermacher
    chat in the terminal
    llama cli -hf mradermacher/phi-1-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf mradermacher/phi-1-GGUF:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run Phi-1 (1.3B) · needs ~0.9 GB · from huggingface.co/mradermacher

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/mradermacher/phi-1-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

version

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Phi-1 (1.3B) · needs ~0.9 GB · from huggingface.co/mradermacher
    chat in the terminal
    ollama run hf.co/mradermacher/phi-1-GGUF:Q4_K_M

Other ways to run it

  1. Run it as published · needs ~2.84 GB · from huggingface.co/microsoft (official)
    uv run --with torch --with 'transformers<5' python -c "from transformers import pipeline; print(pipeline('text-generation', model='microsoft/phi-1')('def fibonacci(n):', max_new_tokens=40)[0]['generated_text'])"

Official weights and documentation: huggingface.co/microsoft ↗

$ how the score was worked out · CRUXEval published evaluation… 76%, Open Portuguese LLM Leaderboa… 16%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
CRUXEval published evaluation results (authors' generations and scores)2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 76%MIT
Open Portuguese LLM Leaderboard (archived)1▇▇▇▇▇ 16%Apache-2.0
prior · pull toward the middle▇▇ 8%
weighted average= 20.7

not counted (2)

Textbooks Are All You Need — authors' own runsHumanEval [correct]50.6%fewer than three models have this test yet, or this source ran it for this model alone
Textbooks Are All You Need — authors' own runsMBPP [correct]55.5%fewer than three models have this test yet, or this source ran it for this model alone

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json