Skip to content
Moojaz — موجز
Search
Menu
HomeRegionsMaghrebAfricaMiddle EastEuropeAsiaAmericasWorldSectionsPoliticsEconomySecurityHealthScienceTechnologyAIEnvironmentCultureSportsSearch
United States of America
File

ServiceNow: Apriel speech models

This dossier tracks ServiceNow’s Apriel models, speech functions, licences and deployment limits.

Last updated : 1 key moments
ServiceNow logo identifying the model’s developer.
ServiceNow logo identifying the model’s developer. — ServiceNow / VulcanSphere / Wikimedia Commons — public domain, PD-textlogo ; PNG rendering.

Where things stand

As of 29 October 2026, Apriel Voice 27B Preview is available under MIT, without native speech synthesis.

Reported information

ServiceNow opens Apriel Voice 27B, which hears speech but answers in text

ServiceNow published Apriel Voice 27B Preview on Hugging Face under the MIT licence on 29 October 2026. The model understands spoken requests and produces text responses and tool calls; it does not generate spoken replies itself.

ServiceNow logo identifying the model’s developer.
ServiceNow logo identifying the model’s developer. — ServiceNow / VulcanSphere / Wikimedia Commons — public domain, PD-textlogo ; PNG rendering.

It combines a Whisper encoder with a Gemma 4 backbone and retains text and image inputs. A speaking assistant needs an external speech synthesiser and turn-handling logic.

The model card warns of uneven performance across languages, problems with noise and long recordings, and potentially incorrect tool calls. Team-reported scores do not guarantee the same performance in another application.

Sources 2