[whatmodel.sh](../../README.md) / Granite 3.1 8B Instruct
# Granite 3.1 8B Instruct
IBM · released 2024-12 · 8.17B params · apache-2.0
**score 44.4** ±0.9 · #489 of 732 · from 19 benchmark results across 6 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Granite 3.1 8B Instruct among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃●█▄▅▅▄▄▂▁ #270 of 313* 3.2% 1
chat & writing ▁▁▁▃▆●▆█▆▇▇▃ #305 of 402 18% 3
math ▁▁▃▅●▆▄█▄▂▁▁ #315 of 431 33% 3
agent tasks · #270 of 313 · 3.2% of its score
Beats [DeepSeek R1 0528 Qwen3 8B](deepseek-r1-0528-qwen3-8b.md), [Magistral Small 1.0 (2506)](magistral-small-2506.md), [Llama 4 Scout (109B-A17B)](llama-4-scout.md) here, though they rank higher overall. Trails [Granite 4.0 H 1B](granite-4-0-h-1b.md), [Granite 4.0 1B](granite-4-0-1b.md), [SmolLM3 3B](smollm3-3b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
BFCL V4 (December 2025 archive) 27.1% ▁▁▁▂●█▅▅▆▂▂▂ beats 20/77 hard · avg 31% 3.2% typical
chat & writing · #305 of 402 · 18% of its score
Beats [Ministral 3 14B Reasoning 2512](ministral-3-14b-reasoning.md), [Gemma 4 E4B](gemma-4-e4b.md), [OpenThinker3 7B](openthinker3-7b.md) here, though they rank higher overall. Trails [LFM2 350M](lfm2-350m.md), [EXAONE 3.5 2.4B Instruct](exaone-3-5-2-4b.md), [LFM2 700M](lfm2-700m.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
LMArena (text) 1208.6 ▁▁▁▂▃●▃▆▆▇█▃ beats 72/385 avg 1260 9.1% blunt
IFEval (aggregation unspecified) … 73.2% ▄▄·▄······●█ beats 3/5 easy · avg 78% 5.6% at its level · several runs
Arena-Hard (version/judge unspeci… 37.6% █·▄··●·····█ beats 3/5 medium · avg 55% 3.6% at its level
math · #315 of 431 · 33% of its score
Beats [Molmo 2 8B](molmo-2-8b.md), [GPT-4 (0613)](gpt-4-0613.md), [Athene-70B](athene-70b.md) here, though they rank higher overall. Trails [DeepSeek R1 Distill Qwen 1.5B](deepseek-r1-distill-qwen-1-5b.md), [Tulu 2 DPO 7B](tulu-2-dpo-7b.md), [Gemma 7B (base)](gemma-7b-base.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
GSM8K (maker) 79.1% █··▄·····●·█ beats 3/5 very easy · avg 89% 23% sharp · several runs
AIME 2024 (maker) 2.0% █···●▄▄····▄ beats 2/5 very hard · avg 8.9% 6.7% sharp · too hard for it
MATH-500 (maker) 48.7% █···●·▄▄···▄ beats 2/5 easy · avg 70% 4.2% at its level
other tests (9)
test result vs the others in its run difficulty counts why
MMLU (maker) 66.8% ▄█········▄● beats 4/5 easy · avg 76% 12% at its level · several runs
BIG-Bench Hard (maker) 69.9% █·▄·····▄●·▄ beats 4/5 easy · avg 81% 10% sharp · at its level
ARC Challenge (maker) 62.6% ▄▁▁▂▁▄█▇●▂▃▂ beats 109/140 easy · avg 77% 8.5% sharp · at its level
WinoGrande (maker) 75.4% ▁▁▁▁▂▁▁▅▇█●▁ beats 114/131 medium · avg 80% 4.2% at its level
PopQA (maker) 28.7% ▄·█······▄·● beats 5/5 hard · avg 33% 3.6% at its level
HellaSwag (maker) 84.5% ▁▁▁▂▂▂▄▆█▇●▂ beats 125/135 easy · avg 87% 2.7% typical
Open PT-LLM Leaderboard average (… 72.3% ▃▁▁▁▁▁▁▁▂●▅▁ beats 195/276 medium · avg 72% 2.2% at its level · noisy
Vectara hallucination leaderboard… 91.4% ▁▁▂▃●▄█▆▃▂▂▁ beats 30/140 very easy · avg 94% 0.7% blunt · noisy
DROP (maker) 58.6% ▃▃·····▃···● beats 4/5 medium · avg 54% 0.4% blunt · at its level
source results share of the score
ibm-granite/granite-3.3-8b-instruct mod… 9 53% ▇▇▇▇▇▇
ibm-granite/granite-3.1-3b-a800m-instru… 6 31% ▇▇▇▇
LMArena leaderboard dataset 1 9% ▇
Berkeley Function Calling Leaderboard r… 1 3% ▇
Open Portuguese LLM Leaderboard 1 2% ▇
Vectara hallucination leaderboard 1 1% ▇
prior: pull toward the middle 0% ▇
weighted average = 44.4
Licences: ibm-granite/granite-3.3-8b-instruct model card ([Apache-2.0](../../licenses/ibm-granite--granite-3.3-8b-instruct.txt)) · ibm-granite/granite-3.1-3b-a800m-instruct model card ([Apache-2.0](https://huggingface.co/ibm-granite/granite-3.1-3b-a800m-instruct/blob/a02780686e08a03fe0d2679a293b5c74a90efa89/README.md)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Berkeley Function Calling Leaderboard result archive ([Apache-2.0](../../licenses/bfcl-result-archive-v4-Apache-2.0.txt)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Vectara hallucination leaderboard ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
## How to run it
### llama.cpp (recommended)
1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md)
```sh
# macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
# Windows
irm https://llama.app/install.ps1 | iex
```
2. Run Granite 3.1 8B Instruct (Q4_K_M, needs ~4.9 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/granite-3.1-8b-instruct-GGUF/tree/c7cc690391c3512b3af7fdb9362aa8fbed4f6050))
```sh
llama cli -hf lmstudio-community/granite-3.1-8b-instruct-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf lmstudio-community/granite-3.1-8b-instruct-GGUF:Q4_K_M
```
Other versions: Q8_0 (~8.7 GB) `llama cli -hf lmstudio-community/granite-3.1-8b-instruct-GGUF:Q8_0`
### LM Studio
1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download)
2. Run Granite 3.1 8B Instruct (Q4_K_M, needs ~4.9 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/granite-3.1-8b-instruct-GGUF/tree/c7cc690391c3512b3af7fdb9362aa8fbed4f6050))
```sh
lms get https://huggingface.co/lmstudio-community/granite-3.1-8b-instruct-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q8_0 (~8.7 GB) `lms get https://huggingface.co/lmstudio-community/granite-3.1-8b-instruct-GGUF@Q8_0 -y`
### Ollama
1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download)
```sh
# Linux
curl -fsSL https://ollama.com/install.sh | sh
```
2. Run Granite 3.1 8B Instruct (Q4_K_M, needs ~5.4 GB, from [ollama.com](https://ollama.com/library/granite3.1-dense/tags))
```sh
ollama run granite3.1-dense:8b-instruct-q4_K_M
```
Other versions: Q8_0 (~10.3 GB) `ollama run granite3.1-dense:8b-instruct-q8_0`
Model repository and documentation: [huggingface.co/ibm-granite](https://huggingface.co/ibm-granite/granite-3.1-8b-instruct)
[interactive version](https://whatmodel.sh/model/granite-3-1-8b) · [JSON](https://whatmodel.sh/api?format=json&all=1)