whatmodel.sh / StarCoder 15.5B

StarCoder 15.5B

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

StarCoder 15.5B is an open-weights language model from BigCode, released April 2023. Its published size is 15.82B parameters. Licence: bigcode-openrail-m.

Catalogue identity notes

Comparison-only. Maker repo bigcode/starcoder @827cd77: licence tag bigcode-openrail-m. 15,819,446,272 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2023-04: month the maker's Hugging Face repo was created (createdAt 2023-04-24). StarCoder = StarCoderBase further trained on Python; not StarCoderBase and not StarCoder2.

34.5±0.7#606 of 719 LLMsfrom 5 benchmark results across 2 sources

$ how it compares

The LLMs ranked either side of StarCoder 15.5B, out of 828 with a score, then more from BigCode.

#modelscorewhereRAMreleased
699Granite 3.0 3B-A800M Instruct IBM35±1local~3.7GB2024-10
700DeepSeek LLM 7B Chat DeepSeek35±2local~5.7GB2023-11
701StarCoder 15.5B BigCode35±1local~11GB2023-04
702Granite 3.1 3B-A800M Instruct IBM35±2local~3.6GB2024-12
703Llama 2 13B (base) Meta34±1local~9.3GB2023-07
more from BigCode
521StarCoder2-15B-Instruct-v0.1 BigCode46±1local~11GB2024-04
639StarCoder2 15B BigCode39±1local~11GB2024-02
698StarCoder2 7B BigCode35±1local~6.1GB2024-02

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▆██▇▆▄▃▁ #415 of 458 · 90% of its score · 3 tests

Beats DeepSeek R1 Distill Qwen 7B, Qwen3.5 2B, DeepSeek R1 Distill Llama 8B here, though they rank higher overall. Trails Llama 3.2 1B Instruct, SmolLM2 1.7B Instruct, Granite 3B Code Base (2K) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEval +134.1% ✎▁·▁▁▃▄▆▅█▃▁▁ beats 19/99easy · avg 79%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 59%sharp · at its level · several runs
HumanEval+29.3% ✎▁·▁▁▁▃▄▄▄█▃▁ beats 19/99easy · avg 74%▇▇▇▇▇▇▇▇▇▇ 26%sharp · at its level
MBPP+46.1% ✎▁···▁▁▂▅▇█▆▂ beats 13/75easy · avg 68%▇▇ 5.3%at its level
other tests1 not in a task · 10% of its score
testresultvs the others in its rundifficultycountswhy
EvoEval (overall)23.5% ✎▂▂▃▄▄█▁▄▅▃▁▃ beats 8/42medium · avg 55%▇▇▇▇ 10%typical

$ 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 StarCoder 15.5B · needs ~10.1 GB · from huggingface.co/mradermacher
    chat in the terminal
    llama cli -hf mradermacher/starcoder-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf mradermacher/starcoder-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 StarCoder 15.5B · needs ~10.1 GB · from huggingface.co/mradermacher

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/mradermacher/starcoder-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 StarCoder 15.5B · needs ~9 GB · from ollama.com
    chat in the terminal
    ollama run starcoder:15b

Other ways to run it

  1. Run it as published · needs ~63.28 GB · from huggingface.co/bigcode (official)
    uv run --with torch --with 'transformers<5' --with accelerate python -c "from transformers import pipeline; print(pipeline('text-generation', model='bigcode/starcoder', torch_dtype='auto', device_map='auto')('Once upon a time', max_new_tokens=40)[0]['generated_text'])"

Official weights and documentation: huggingface.co/bigcode ↗

$ how the score was worked out · EvalPlus / EvalPerf published… 70%, Top Leaderboard Ranking = Top… 30%

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
EvalPlus / EvalPerf published evaluation runs3▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 70%Apache-2.0
Top Leaderboard Ranking = Top Coding Proficiency, Always? EvoEval: Evolving Coding Benchmarks via LLM — authors' own runs2▇▇▇▇▇▇▇▇▇ 30%CC BY 4.0
prior · pull toward the middle▇ 0%
weighted average= 34.5

$ 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