BioGPU tech stack

Every layer AI needs to work on cellular biology, built as one system.

BioGPU is the Cellular Biology Runtime for AI Agents. It runs from the agent that designs the experiment to the tokens that bring the result back into AI. Anima built every layer, and each one depends on the layer before it.

One integrated stack · design goes down, evidence comes back up

The components

Every cycle takes a network hypothesis into cells and brings back the resolved network.

  1. 01

    BioGPU Agent

    Anima's specialized agent and its expert in cellular network experiments. Given a disease, a target and a question, it proposes the network hypothesis, designs the cellular network experiments, chooses the molecular-specific visualizers and perturbations, and decides the next experiment from what comes back. It carries more than a decade of Anima's experience in designing these experiments.

    Designs the experiment · chooses the visualizers · directs the next step

  2. 02

    Cellular network experiments

    The wet-lab layer. Experiments run in patient-derived and other disease-relevant cells, measuring many molecular activities in parallel across conditions, perturbations and timepoints, and imaging every cell.

    1,000+ experiments per run · 150+ cellular systems run on BioGPU

    Explore cellular systems →
  3. 03

    Molecular-specific visualizers

    Each visualizer makes one molecular activity visible inside the cell: a protein interaction, a translocation, translation, RNA localization or a signaling state. Placed across the network, they show how the activities change together.

    1,000+ cellular activities with their own visualizers

  4. 04

    Visual Biology Foundation Model

    The model recognizes the molecular activities in the experimental images, at single-cell resolution, and derives the state of the network. It recognizes; it does not predict.

    Trained on 2B+ proprietary experimental images

  5. 05

    Visual Biology Tokens

    What the model recognizes is returned as Visual Biology Tokens. Each token links three parts: the image evidence, the biological meaning and the experimental context. They go straight into an AI agent's context window.

    Image evidence · biological meaning · experimental context

  6. 06

    AI agents

    Your AI agents, connected through BioGPU Runtime Skills, and the BioGPU Agent reason across the tokens, refine the network and ask for the next run. Drug-design AI can then read every molecule it proposes against the network.

    Reason · Run · Recognize · Resolve

  7. 07

    BioGPU Program Runtime

    For each target program, the stack is set up as a dedicated runtime: the target, a disease-relevant cellular system and the cellular network resolved around it. Every compound in the program then runs on it.

    One runtime per target program

    See how a program runs →

The core of the stack · Visual Biology Foundation Model

The model sees the whole network.

At the core of BioGPU, the Visual Biology Foundation Model sees what is happening across the whole cellular network, in disease-relevant cells, and hands that picture to AI agents.

Explore the Visual Biology Foundation Model →

Illustrative · cell images organized by the visual patterns the model has learned
2B+Experimental images
1,000+Cellular activities
150+Cellular systems
  1. 01 · Visual languageTeaching AI the visual language of molecular biology.Molecular activities leave recognizable patterns. The model learns them from the images.
  2. 02 · Proprietary dataA foundation built on proprietary experimental biology.2B+ experimental images across 1,000+ cellular activities and 150+ cellular systems.
  3. 03 · GeneralizationRecognizing biology beyond the training data.It places proteins and phenotypes it never saw near the visual patterns it learned.
  4. 04 · Biological evidenceTurning image recognition into biological evidence.Every recognized pattern stays linked to the experiment that produced it.
  5. 05 · Network stateFrom single activities to the state of the whole network.Activities recognized across the network become its state, ready for AI agents to reason on.
  6. 06 · Single cellEvery cell read on its own.Responses are recognized cell by cell, so a small population that responds differently stays visible.

Reasoning on the network

With the whole network in view, AI designs better compounds and sharper hypotheses.

Each finding adds to one picture of the disease network. With that picture in view, AI agents see connections no single experiment reveals, and use them for two things: designing the next AI compound, or forming the next network hypothesis to test.

Illustrative · tokens on the network, the agent's context, and the next hypothesis or compound design

Because Anima built every layer, each round sharpens the next: the agent learns what to run, and the model learns to recognize more.

Give your AI the biology it is missing.

Pharma team? See Pharma Workflows →

Partner with us [email protected]