Proton launches AI email writing assistant that can run locally or on the server

Proton The local time yesterday launched AI EmailWriting Assistant Proton Scribe. This auxiliary software is built into the Proton Mail mailbox, which can save time in writing emails and improve the productivity of corporate teams.

Proton launches AI email writing assistant that can run locally or on the server

Proton Scribe Supports generating email drafts from scratch based on user needs, can also be used to proofread, shorten or formalize existing drafts of text.

Proton says that Proton Scribe Assistant is designed to protect data privacy and can be run on a dedicated server like other similar software.Runs on user hardware.

The latter method canEnsure that user data does not leave the local, more suitable for processing sensitive data or scenarios where network threats may exist.

For macOS users, if you want to use Proton Scribe natively, you need an M-series chip and 8GB of memory. It can be used in the Chromium-based browser and the Proton Mail desktop application.

For Windows or Linux system users, IT Home summarizes the requirements for local use as follows:

  • CPU: At least a 7th generation Intel Core processor or AMD Ryzen 1000 series processor is required;

  • GPU: At least 6GB video memoryAnd support independent graphics cards with FP16 data format;

  • Memory: At least 8GB RAM;

  • Supported platforms: Only available on Chromium-based browsers.

Although the Proton Scribe assistant has the potential to generate text in other languages, Proton officials say that itCurrently only supports English.

statement:The content is collected from various media platforms such as public websites. If the included content infringes on your rights, please contact us by email and we will deal with it as soon as possible.
Information

Microsoft Azure AI officially launches GPT-4o mini, which is cheaper and faster

2024-7-19 9:33:57

Information

SenseTime Jueying launches the industry's first native multi-modal large model vehicle-side deployment: 8 billion parameters, 40 tokens per second

2024-7-20 8:21:33

Search