London-based consumer technology company Nothing has introduced a new artificial intelligence powered dictation tool for its devices. The feature, which operates entirely on-device, offers support for over 100 languages and does not require an internet connection to function.
The announcement was made by Nothing on its official channels. The company stated that the tool is designed to process speech into text locally on the device, a move aimed at addressing user privacy concerns. By handling the processing onboard rather than in the cloud, the company seeks to reduce the risk of audio data being transmitted to external servers.
According to Nothing, the on-device processing capability is a key differentiator. The tool is expected to be integrated into the company’s smartphone operating system, Nothing OS. This integration will allow users to access dictation features across various applications, including messaging and note taking, without needing to open a separate application.
The decision to support over 100 languages is intended to broaden the tool’s usability for an international customer base. Nothing has a global user community, and the multilingual support aligns with its strategy to serve markets beyond its initial launch regions. The company has not yet published a full list of the included languages but has confirmed that major global languages are covered.
Privacy and performance considerations
The company emphasized that the tool’s offline functionality is significant for both privacy and performance. On-device processing can reduce latency, as the data does not need to travel to a remote server and back. This can result in faster transcription speeds, especially in areas with poor network connectivity.
From a privacy standpoint, keeping the dictation process local means that the user’s spoken words are not stored or analyzed by third party servers. Nothing has previously highlighted user data protection as a priority, and this feature reinforces that commitment.
Broader industry context
The launch follows a wider trend in the technology sector where major companies are investing in local AI processing. Apple, Google, and Samsung have all developed on-device voice recognition and transcription features. Nothing’s entry into this space with a focus on broad language support positions it competitively within the mid range smartphone market.
The feature is expected to roll out via a software update. Nothing has not specified a precise release date but indicated that it will be available to users in the coming weeks.
Additional technical details
The dictation tool utilizes machine learning models that have been optimized for mobile hardware. These models are designed to run efficiently on the Qualcomm and MediaTek processors used in Nothing’s current handset lineup. The company has not disclosed the specific model size or memory requirements.
Early demonstrations provided by the company show the tool can handle continuous speech and punctuation. It also supports multiple languages within a single dictation session, allowing bilingual users to switch between languages without manually changing settings.
Potential limitations and future updates
While the tool represents a notable addition, it may have limitations compared to cloud based alternatives. Cloud services can leverage larger language models and more frequent updates. Nothing has stated that it plans to refine the tool over time based on user feedback and to improve accuracy for less common languages.
The feature is currently limited to Nothing’s own devices, including the Phone (1), Phone (2), and the upcoming Phone (3). There is no indication that it will be made available to other Android manufacturers.
Source: Nothing
