Granite Switch 4.1: chat, decisions and speech from one endpoint
On October 15, 2026, IBM published Granite Switch 4.1, in 8B and 30B versions, combining chat, decision and safety adapters and a speech recognition model while using about the same GPU memory as the base model.
Source: Granite Switch 4.1: Chat, Decisions, and Speech from One Endpoint (huggingface.co). Text prepared with AI from this source.
What happened and what to do
According to IBM, Granite-Switch-4.1-8B and -30B include a general-purpose decision adapter, a Guardian adapter for safety decisions and Granite-Speech-5.0-Turbo-CTC, which the company describes as the fastest speech recognition model on the OpenASR leaderboard. The adapters reuse the base model's cache, so one endpoint switches between chat, classification and audio.
For a company, this may mean fewer GPUs to serve voice, content screening and support. Before adopting it, test with real audio and text from your domain and language, since IBM says more data will be published in the coming weeks.
How the consultancy can help
Wendelmaques can assess whether Granite Switch fits your case, set up a test with your own data and, if it makes sense, run the model on your infrastructure with monitoring. The service is on demand and the scope is defined with your team.
Next step
If your team handles voice or needs to classify content, send a short description of the case and receive a scoped proposal.
Let's assess your project
I reply with scope and a proposal.