granite-speech-3.3-2b is part of IBM's Granite family, a 2B parameter model supporting speech and text modalities. It includes built-in tool calling, system prompts, and is optimized for helpful assistant behavior. Available on Hugging Face, it is aimed at developers creating voice-enabled applications or multimodal AI agents that require efficient on-device or cloud inference.
In the Voice, TTS & speech space, Granite Speech 3.3 2b takes a focused approach. It focuses on integrating speech understanding and conversational AI capabilities with tool use in a single compact model. It is built as an open-source project for developers. Granite Speech 3.3 2b is open source under the Open Source license. Granite Speech 3.3 2b is available on the web, the command line, and API.
Behind Granite Speech 3.3 2b is IBM, based in the United States, and the product first shipped in 2025. Development happens publicly on GitHub with 45 stars and 2 commits in the last 90 days. It operates in a well-populated space: PulseGate tracks 13 similar tools. Key capabilities include Speech Processing, Tool Calling, and System Prompts. It exposes integrations via a public API.
Latest indexed changes and source events
ibm-granite/granite-speech-3.3-2b verified by the PulseGate indexer
Other apps tracked under the same category.
