Model overview
What this model is for
granite3.3:2b is a compact IBM Granite language model intended for reasoning and instruction-following tasks, with tool-calling support. Ollama describes the Granite 2B and 8B models as having an advertised 128K-token context length, which can be useful for processing long documents, extended conversations, or multi-step prompts. For developers, the model is a candidate for local reasoning workflows and applications that need to invoke tools, provided the surrounding runtime supports tool calling. The supplied material does not identify any supported input or output modalities, so multimodal use should not be assumed. It also does not provide an inference-speed benchmark, runtime context configuration, pricing, or a detailed evaluation of accuracy, latency, or resource requirements. The page identifies the model as granite3.3:2b and describes fine-tuning for improved reasoning and instruction following, but does not include implementation guidance or task-specific benchmarks. Licensing terms are not available in the supplied material and should be verified before redistribution or commercial deployment.