Skip to main content
Ollama ExplorerBeta
Back to models
localtools

granite3.2

Granite-3.2 is a family of long-context AI models from IBM Granite fine-tuned for thinking capabilities.

467K pulls9 tagsSource updated Feb 26, 2025128K max context

Quick start

ollama run granite3.2:latest

Runnable variants

Memory is a rough file-size estimate; context length, runtime and GPU offload change it.

9 of 9 matching variants shown.

2bTypelocalInputstextDigest9d79a41f2f75Download1.5 GBParameters2bQuantizationUnknownContext128KEst. memory≈ 3 GB
ollama run granite3.2:2b
2b-instruct-fp16TypelocalInputstextDigestc86bd0b2f0bdDownload5.1 GBParameters2bQuantizationFP16Context128KEst. memory≈ 7 GB
ollama run granite3.2:2b-instruct-fp16
2b-instruct-q4_K_MTypelocalInputstextDigest9d79a41f2f75Download1.5 GBParameters2bQuantizationQ4_K_MContext128KEst. memory≈ 3 GB
ollama run granite3.2:2b-instruct-q4_K_M
2b-instruct-q8_0TypelocalInputstextDigestf6f4d0d4b4faDownload2.7 GBParameters2bQuantizationQ8_0Context128KEst. memory≈ 4.5 GB
ollama run granite3.2:2b-instruct-q8_0
8bTypelocalInputstextDigest9bcb3335083fDownload4.9 GBParameters8bQuantizationUnknownContext128KEst. memory≈ 6.5 GB
ollama run granite3.2:8b
8b-instruct-fp16TypelocalInputstextDigestb9679f6ae759Download16 GBParameters8bQuantizationFP16Context128KEst. memory≈ 19 GB
ollama run granite3.2:8b-instruct-fp16
8b-instruct-q4_K_MTypelocalInputstextDigest9bcb3335083fDownload4.9 GBParameters8bQuantizationQ4_K_MContext128KEst. memory≈ 6.5 GB
ollama run granite3.2:8b-instruct-q4_K_M
8b-instruct-q8_0TypelocalInputstextDigest1b54499df0ffDownload8.7 GBParameters8bQuantizationQ8_0Context128KEst. memory≈ 11 GB
ollama run granite3.2:8b-instruct-q8_0
latestdefaultTypelocalInputstextDigest9bcb3335083fDownload4.9 GBParametersUnknownQuantizationUnknownContext128KEst. memory≈ 6.5 GB
ollama run granite3.2:latest

Source-listed applications

Claude CodeHermes AgentOpenClawOpenCode

Related models

granite3.3local

IBM Granite 2B and 8B models are 128K context length language models that have been fine-tuned for improved reasoning and instruction-following capabilities.

hermes3local

Hermes 3 is the latest version of the flagship Hermes series of LLMs by Nous Research

granite4.1local

IBM Granite Models are a family of enterprise-ready, open foundation models that support multilingual capabilities, coding, retrieval-augmented generation (RAG), tool use, and structured JSON output. Released under Apache 2.0 license.

glm-ocrlocal

GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture.