whatmodel.sh / Granite 20B Code Base (8K)

Granite 20B Code Base (8K)

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

Granite 20B Code Base (8K) is an open-weights language model from IBM, released April 2024. Its published size is 20.07B parameters. Licence: apache-2.0.

Catalogue identity notes

Comparison-only. Maker repo ibm-granite/granite-20b-code-base-8k @8bbfb1d: licence tag apache-2.0. 20,066,993,152 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2024-04: month the maker's Hugging Face repo was created (createdAt 2024-04-21). May 2024 Granite Code release; not r1.1.

37.0±1.2#577 of 719 LLMsfrom 7 benchmark results across 2 sources

$ how it compares

The LLMs ranked either side of Granite 20B Code Base (8K), out of 828 with a score, then more from IBM.

#modelscorewhereRAMreleased
668Gemma 2 2B IT Google37±1local~3.1GB2024-07
669Qwen2.5 Coder 0.5B Instruct Alibaba37±1local~1.8GB2024-11
670Granite 20B Code Base (8K) IBM37±1local~15GB2024-04
671Gemma 1.1 7B IT Google37±1local~7.1GB2024-04
672Code Llama 7B Python Meta37±1local~5.8GB2023-08
more from IBM
192Granite 4.2 30B IBM63±1api $0.28/M +local~19GB2026-08
239Granite 4.2 8B IBM60±1api $0.11/M +local~6.6GB2026-08
276Granite 4.2 3B IBM57±1api $0.052/M +local~3.7GB2026-08

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▆██▇▆▄▃▁ #396 of 458 · 85% of its score · 6 tests

Beats DeepSeek R1 Distill Qwen 7B, Qwen3.5 2B, DeepSeek R1 Distill Llama 8B here, though they rank higher overall. Trails Qwen2.5 1.5B Instruct, WizardLM 13B V1.2, EXAONE 3.5 2.4B Instruct here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEvalSynthesize (HumanEvalPack, 6-language average) lab44.3% ✎▁·▂▁▂▆▅█▄▄▁▂ beats 25/35medium · avg 64%▇▇▇▇▇▇▇▇▇▇ 25%at its level
HumanEvalExplain (HumanEvalPack, 6-language average) lab18.8% ✎▄▇█▃▄▇▄▁▇▃·▁ beats 15/35medium · avg 49%▇▇▇▇▇▇ 14%typical
MBPP lab43.8% ✎█·▅▃▃▃▅▃▃▅·▅ beats 10/15easy · avg 70%▇▇▇▇▇▇ 14%at its level
HumanEvalFix (HumanEvalPack, 6-language average) lab17.7% ✎▁·▁▁▄██▄█▁▅▅ beats 17/35medium · avg 56%▇▇▇▇▇ 12%sharp
CRUXEval-I lab39.1% ✎▃▅█▃▃▅·▃▃▅·▅ beats 10/15medium · avg 59%▇▇▇▇ 11%at its level
CRUXEval-O lab37.5% ✎▃·▃██▅▃·▃▃▅▃ beats 10/15medium · avg 55%▇▇▇▇ 8.9%blunt · at its level
other tests1 not in a task · 15% of its score
testresultvs the others in its rundifficultycountswhy
BigCodeBench (Complete)25.8% ✎▂▁▃▄▂▅▇█▆▂▂▂ beats 5/58medium · avg 50%▇▇▇▇▇▇ 15%at its level

$ where to run it

llama.cpp · recommended

  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 Granite 20B Code Base (8K) · needs ~12.82 GB · from huggingface.co/ibm-granite
    chat in the terminal
    llama cli -hf ibm-granite/granite-20b-code-base-8k-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf ibm-granite/granite-20b-code-base-8k-GGUF:Q4_K_M

LM Studio

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

    download it (macOS / Windows / Linux) ↗

  2. Run Granite 20B Code Base (8K) · needs ~12.82 GB · from huggingface.co/ibm-granite

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/ibm-granite/granite-20b-code-base-8k-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Granite 20B Code Base (8K) · needs ~12.82 GB · from huggingface.co/ibm-granite
    chat in the terminal
    ollama run hf.co/ibm-granite/granite-20b-code-base-8k-GGUF:Q4_K_M

Weights and documentation: huggingface.co/ibm-granite ↗

$ how the score was worked out · Granite Code Models: A Family… 85%, BigCodeBench paper (Zhuo et a… 15%

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
Granite Code Models: A Family of Open Foundation Models for Code Intelligence — authors' own runs · lab-reported6▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 85%CC BY 4.0
BigCodeBench paper (Zhuo et al., ICLR 2025), arXiv 2406.15877v41▇▇▇▇ 15%CC BY 4.0
prior · pull toward the middle▇ 1%
weighted average= 37.0

$ 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