Skip to content
Eugeny

Capabilities

One family, every sense.

Our European release brings language, vision, image, video and voice into one system, with Soffio for knowledge and Matchbox for access. One model family that sees, listens, reads, creates and remembers.

The model behind it

Illustration: a woman reads a book while sitting on a stone wall beneath an olive tree, above a coastal European town with a domed church, cypresses, hills and sailboats.

Vision

Vision that does the work.

Eugeny reads a photo, a scan, a shelf, a page or a frame as a task, not a caption.

Images open new conversations.

Bring visual context into the same conversation as your ideas.

  • Describe what you see
  • Ask questions
  • Explore different views
  • Turn ideas into images
Eugeny

What can you tell me about this?

Illustration of a reader overlooking a coastal town

A reader beneath an olive tree, looking out over a town by the sea.

A new perspective, whenever you need one.

Understand images

Read a scene, a product, a room or a page and reason about it in the same conversation.

Detect

Find what matters and where it is: objects, labels, defects, faces, the parts of a frame that change a decision.

Segment

Lift a subject from its background or a product from a shelf, ready to isolate, measure or edit.

Count

Count what is there, and show the count with the objects behind it.

Read documents

Pull text and structure from pages, scans and forms, including the fields that only make sense together.

Describe

Put what is visible into clear language at the length the work needs.

Media

Images, video, voice.

The media models read the language model's conditioning, so they answer the request you actually made.

  • Image

    Generate and edit product shots, scenes and graphics, with references that preserve product identity.

  • Video

    Generate and edit footage, keeping a subject, a product or a performance consistent across shots.

  • Voice

    Turn a script into a performance with the delivery the piece needs.

Different ways to explore.

Move between words, images, video and voice without losing the thread.

Text

Turn thinking into progress.

  • A thought.
  • A question.
  • A few words.
  • A clearer idea.

Write, think and create.

Image

See a new perspective.

Illustration of a coastal town

Explore and create images.

Video

Bring ideas to life.

Illustration of a woman working beside a window

Work across moments and scenes.

Voice

A natural way to think.

Talk it through.

Upscaling

Small sources, print-ready results.

Upscaling for the material real work arrives in.

A phone crop, a frame pulled from video, a catalogue scan from years ago. Eugeny restores and enlarges it for packaging, print and large screens, holding structure and lettering together instead of smoothing them into a blur.

Soffio

Knowledge, memory, team.

Soffio connects your team in a peer-to-peer network, sharing the skills and memories you choose.

Knowledge

Notes, documents and decisions in one governed place, retrieved when they matter.

Memory

What Eugeny remembers between sessions: explicit, scoped, and removable.

Team

Shared knowledge that belongs to the team that shared it, with an owner and a boundary.

Access

Connected, without giving up the keys.

Matchbox opens the systems a company already runs.

Repositories, hosting, mail, files, ad accounts, servers. Matchbox connects each task to the access it needs. It holds provider credentials in the broker and gives your team one place to manage permissions.

Finetuning

Make it yours.

Local-first finetuning, in the app where the work happens.

Finetune Eugeny on your own material and it learns your vocabulary, your products and the calls your work turns on, inside your own deployment. The native app brings conversation, files, images, video, voice and team knowledge together in one place.

How the model is built is on the model page.