whatmodel.sh / Granite 3B Code Instruct (2K)

Granite 3B Code Instruct (2K)

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

Granite 3B Code Instruct (2K) is an open-weights language model from IBM, released April 2024. Its published size is 3.48B parameters. Licence: apache-2.0.

Catalogue identity notes

Comparison-only. Maker repo ibm-granite/granite-3b-code-instruct-2k @7bba6a5: licence tag apache-2.0. 3,482,503,680 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2024-04: month the maker's Hugging Face repo was created (createdAt 2024-04-26). May 2024 Granite Code release; not the 128K variant.

38.0±1.5#564 of 719 LLMsfrom 6 benchmark results across 3 sources

$ how it compares

The LLMs ranked either side of Granite 3B Code Instruct (2K), out of 828 with a score, then more from IBM.

#modelscorewhereRAMreleased
655Apertus 8B Instruct (2509) Swiss AI38±4local~6.3GB2025-08
656LLaMA 65B Meta38±1comparison only2023-02
657Granite 3B Code Instruct (2K) IBM38±2local~3.7GB2024-04
658DeepSeek-V2 Lite Chat (16B-A2.4B) DeepSeek38±1local~11GB2024-05
659BioMistral-7B BioMistral38±3local~6.1GB2024-02
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▁▁▁▃▆██▇▆▄▃▁ #373 of 458 · 73% of its score · 3 tests

Beats LFM2.5 8B-A1B, DeepSeek R1 Distill Qwen 7B, Qwen3.5 2B here, though they rank higher overall. Trails Qwen2.5 1.5B Instruct, WizardLM 13B V1.2 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEvalSynthesize (HumanEvalPack, 6-language average) lab39.6% ✎▁·▂▁▂▆▅█▄▄▁▂ beats 18/35medium · avg 64%▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 34%at its level
HumanEvalExplain (HumanEvalPack, 6-language average) lab26.0% ✎▄▇█▃▄▇▄▁▇▃·▁ beats 21/35medium · avg 49%▇▇▇▇▇▇▇▇ 20%typical
HumanEvalFix (HumanEvalPack, 6-language average) lab27.3% ✎▁·▁▁▄██▄█▁▅▅ beats 22/35medium · avg 56%▇▇▇▇▇▇▇ 18%sharp
other tests3 not in a task · 26% of its score
testresultvs the others in its rundifficultycountswhy
BigCodeBench (Complete)31.5% ✎▂▁▁▄▃▅▇█▆▂▂▂ beats 12/58medium · avg 50%▇▇▇▇▇▇▇▇ 20%at its level
EuroEval Life in the UK (English knowledge MC), test split, 5-shot, accuracy35.5% ✎▁▁▁▁▁▃▁▂▄█▄▃ beats 11/146easy · avg 83%▇ 3.0%at its level · noisy
EuroEval HellaSwag (English), test split, 5-shot generative MC, accuracy28.5% ✎▃·▁▃▂▁▂▅█▇▄▁ beats 11/100medium · avg 68%▇ 2.8%at its level · noisy

$ 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 3B Code Instruct (2K) · needs ~2.13 GB · from huggingface.co/ibm-granite
    chat in the terminal
    llama cli -hf ibm-granite/granite-3b-code-instruct-2k-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-3b-code-instruct-2k-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 3B Code Instruct (2K) · needs ~2.13 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-3b-code-instruct-2k-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 3B Code Instruct (2K) · needs ~2.13 GB · from huggingface.co/ibm-granite
    chat in the terminal
    ollama run hf.co/ibm-granite/granite-3b-code-instruct-2k-GGUF:Q4_K_M

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

$ how the score was worked out · Granite Code Models: A Family… 73%, BigCodeBench paper (Zhuo et a… 20%, EuroEval leaderboards (result… 6%

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-reported3▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 73%CC BY 4.0
BigCodeBench paper (Zhuo et al., ICLR 2025), arXiv 2406.15877v41▇▇▇▇▇▇ 20%CC BY 4.0
EuroEval leaderboards (results collection), English generative leaderboard2▇▇ 6%MIT
prior · pull toward the middle▇ 1%
weighted average= 38.0

$ 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