Techcyte lets labs run in-house AI models in Fusion AP

Techcyte says health systems, academic medical centers and laboratories can now host their own AI models inside Techcyte Fusion AP. The move is meant to help labs move in-house algorithms from development into day-to-day pathology workflows with private hosting, GPU compute and integrated review tools.

Why it matters: - Laboratories building AI can now run their own models inside the same pathology workflow pathologists already use. - Techcyte is aiming to remove the infrastructure burden that often slows in-house AI from testing to deployment. - The feature could help health systems and academic medical centers put local models into use faster while keeping control of hosting and budget.

What happened: - Techcyte announced that health systems, academic medical centers and laboratories can host and run AI models they develop in-house within Techcyte Fusion® AP. - The models are hosted in a private cloud environment provided and managed by Techcyte. - The announcement was made October 9, 2026, from Orem, Utah.

The details: - Development teams package models as standard Docker containers and upload them through an enterprise portal in Fusion. - Techcyte hosts the containers in a GPU-enabled cloud environment. - Results return through Techcyte’s results API. - A Software Developer Kit guides teams through container preparation and integration with Fusion. - Techcyte Flows connect model results to interactive components inside the Fusion case and slide viewer. - Institutions can show heatmaps, quantitative counts and region overlays alongside the slide. - Customer-deployed models use the same user experience as Techcyte’s own algorithms and third-party AI on Fusion AP. - Results appear in the active case workspace, where pathologists review, edit and confirm model output. - Pricing includes a monthly platform fee for managed hosting plus compute credits that are drawn down as models run. - Charges follow actual GPU usage instead of a fixed price per scan. - Credit usage is visible in Fusion. - Institutions can set annual spending caps.

Between the lines: - Techcyte is positioning Fusion AP as a platform for both vendor-built AI and customer-built AI, not just a distribution channel for prepackaged algorithms. - The workflow design suggests the company wants model output to feel native to the pathologist review process rather than like a separate add-on. - The usage-based pricing model may make in-house deployment easier for labs whose model workloads vary over time.

What's next: - At Pathology Visions, Dr. Jansen Seheult, Medical Director of the BloodBytes AI Lab at Mayo Clinic, will discuss LymphoVision, Mayo Clinic’s in-house-developed foundation model. - The discussion will take place during Techcyte’s pre-conference workshop, “How Academic and Private Practice Labs Are Approaching AI-Ready Digital Pathology,” on October 16 at 11 a.m. PT in Seaport Ballroom G at the Manchester Grand Hyatt. - Attendees can also visit Techcyte at booth 410 to learn more about in-house model hosting, Techcyte Flows and AI workflows within Fusion AP. - More information is available at Techcyte’s website.

The bottom line: - Techcyte is giving labs a way to move their own AI models into production inside Fusion AP with hosted infrastructure, integrated workflow tools and consumption-based pricing.