Digital Pathology AI
H

HistAI AI Autopilot

AI Autopilot lets pathologists and researchers train their own AI to classify tissue slides, without writing code. It is made by HistAI, a digital pathology company incorporated in California.

A user connects an AI coding agent such as Claude Code, Codex or Gemini to HistAI's platform and describes the task in plain words. The agent builds a training cohort from more than 66,000 H&E slides whose features were extracted in advance with HistAI's Hibou-L foundation model, trains a classifier, and reports its AUROC, sensitivity, specificity and confusion matrix. The finished model runs on HistAI's cloud platform, CellDX, and Business plan subscribers can download the weights. Pricing is published: plans cost $15 to $39 a month, and the product page puts training at $15 to $50 a model.

The models are not FDA cleared. Certified laboratories may use them in their own laboratory developed tests if they validate them. HistAI keeps ownership of every custom model.

HistAI's main offering is its Data Hub, which licenses 300,000 de identified slides to AI developers. That is a data product, so the grades on this page describe AI Autopilot only.

AI Health Index verifiedNovember 8, 2026
Compare HistAI AI Autopilot with other vendors
Founded
—
Headquarters
California, United States
Categories
pathology-ai
Indexed Products
AI Autopilot, CellDX platform, Hibou foundation models (open source), HistAI Data Hub
Buyer Segments
Pathologists, Pathology laboratories, Academic researchers, Biopharma and AI developers
Assessment

Capability Axes

An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read

AI Capability
AA on AI CentralityThe artificial intelligence is the product. Remove the model and there is nothing left to sell.
Vendor Published

The models are the product. AI Autopilot exists to train and run custom pathology classifiers on whole slide images, built on HistAI's foundation models, and what the user buys is time to train and run models. The viewer and annotation tools on CellDX serve it, and do not stand as the offering. Without the models, nothing is left of this product.

BB on Autonomy and Oversight ModelThe oversight structure is described and one part is missing, commonly the threshold at which the system stops or what happens after it is wrong.
Vendor Published

The user approves each step, and validation is required before clinical use, but no threshold is published. The AI agent asks clarifying questions, the user confirms the cohort and the cost estimate before a job runs, and the agent reports validation figures before the user decides to deploy. HistAI's terms allow a CLIA certified laboratory to use custom model outputs in a laboratory developed test only under the validation requirements they set out. Otherwise the platform is not cleared for clinical use.

What is missing is guidance on when a custom model is good enough: the product page says to ask the agent to improve a model if its metrics are not high enough, with no minimum performance and no statement of how a pathologist reviews outputs.

BB on Model and Technology TransparencyThe approach or the suppliers are named without the version and update discipline behind them. Naming a supplier is the entry to this band both here and on Model Supply Chain Disclosure, which ask different questions of the same disclosure: who receives the data, and what produces the output.
Vendor Published

The foundation model under AI Autopilot is named, without an update practice. HistAI's May 2026 launch post says slides are encoded in advance with Hibou-L, at 1,024 features per tile at 20 times magnification, and classifiers are trained on those features with a multiple instance learning framework and automated hyperparameter search, described in a 2026 preprint. The product page calls the foundation models proprietary, while HistAI released its Hibou models in 2024 under the Apache 2.0 license.

Each job records its configuration, cohort, seed and metric history, but no version or update practice is published for the foundation model or the feature store.

BB on Model Supply Chain DisclosureSubstantial partial disclosure, or a chain that is structurally short: an in house build, a cleared model that cannot be quietly swapped, or a deployment where the transfer does not occur at all. Naming the model provider exits the band below into this one; the axis rises from here on the completeness of the party list and on the terms that govern data once it arrives.
Vendor Published

HistAI builds its own models, and the outside parties in the workflow are named. Custom models are built on its Hibou-L features and run on CellDX, which its launch post says is secured on Microsoft Azure. The work is driven by an AI agent the user chooses, such as Claude Code, Codex or Gemini, using skills HistAI publishes as open source.

The privacy policy names Stripe for payments and refers to cloud computing and analytics providers without naming them. Nothing says what data the user's agent provider receives, and no subprocessor list is published.

CC on Clinical and Operational EvidenceNamed customers, or vendor reported percentages with no method, denominator or reference standard. Scale of use is recorded here and is not treated as evidence of benefit.
Vendor Published

The underlying models have published research behind them; the product's own results have none. HistAI's Hibou foundation models are described in a 2024 preprint on arXiv that reports results on public benchmarks, and a 2026 preprint describes how AI Autopilot trains and deploys classifiers. HistAI shows Stanford, MIT, Harvard, Mayo Clinic and others among the organizations whose researchers use its platform.

AI Autopilot reports validation figures to each user for the model they train. No study of how custom models built this way perform against pathologists or established methods is published, and the preprints are not peer reviewed.

CC on AI Safety and PHI StewardshipGeneral assurances of privacy and security that do not answer the questions artificial intelligence raises: what is retained, what reaches a model, and what happens to it there.
Vendor Published

Under the terms, HistAI holds broad rights to user content, and they do not say whether uploaded slides train its models. The terms (last updated 10 April 2026) give HistAI a perpetual, unrestricted license to content users contribute, and state that every custom model a user trains, with its weights and settings, stays HistAI's exclusive property.

Neither the terms nor the privacy policy says whether uploaded slides or annotations train HistAI's own models or enter its Data Hub. Data is kept as long as necessary, with no period stated. The user's chosen AI agent drives the work through an API, and nothing says what slide data the agent's provider receives. No safety engineering for the models is published.

Regulatory and Compliance
CC on HIPAA and BAA PostureA status is stated but not supported: compliance is claimed, or business associate status stated, without the underlying document, including where the only privacy notice published covers the website rather than the service that handles patients. The mirror case grades here as well: a substantive privacy document that reaches the service, with no statement of business associate status anywhere and no scope position taken. One of the two elements the band above asks for is present, which is a grade below it and not a grade at the floor.
Vendor Published

HIPAA compliance is asserted, without a status or agreement. The privacy policy, effective 22 September 2025, covers the website, the CellDX app and related services, and names HistAI as controller of account data. It says HistAI processes personal and health related data in compliance with HIPAA and the GDPR, and that its slide datasets are de identified. Its terms make the user responsible for complying with HIPAA when uploading slides.

No business associate status is stated and no business associate agreement is published.

DD on Security Certifications and Trust CenterControls are asserted with nothing independent behind them, or nothing is published. Read the note before concluding anything: this is the grade most often corrected on a second pass, because assurance material frequently sits on a parent domain or inside an old announcement rather than on the product pages.
Vendor Published

HistAI publishes no security audit or certification. Its privacy policy describes technical and organizational measures, including encryption of datasets, and its launch post describes Azure managed identities and private storage endpoints, which are the host's controls. No SOC 2 report, ISO/IEC 27001 certificate, HITRUST certification, penetration testing statement or trust center is published.

CC on FDA and Regulatory StatusNo device claim is made and the product is scoped accordingly. Most administrative and operational products sit here and are not penalized for it, because this axis grades the appropriateness of the positioning rather than possession of a clearance.
Vendor Published

HistAI makes no device claim, and scopes the product to research and laboratory developed tests. Its terms (last updated 10 April 2026) state that the platform and custom models are not FDA cleared or FDA approved medical devices.

The terms allow CLIA certified laboratories to use custom model outputs in laboratory developed tests under validation requirements, and the launch post says trained artifacts are licensed for commercial use including laboratory developed test development. That puts regulatory responsibility on the laboratory, under laboratory rules. The FDA's databases list no HistAI device.

CC on AI Governance and Bias DisclosureResponsible artificial intelligence is committed to in policy language with no evaluation behind it. Most of the index sits here.
Vendor Published

Each model's builder sees its validation results, but nothing is published on behavior across groups. The AI agent proposes a training and validation split balanced by class and case, reports AUROC, sensitivity, specificity and a confusion matrix, and each job records its configuration, cohort, seed and metric history. The Hibou preprint reports benchmark results for the foundation models.

No results compared across scanners, stains, organs or patient groups are published for the foundation models under AI Autopilot. No governance framework or monitoring after deployment is described. Custom models trained on small or narrow slide sets are where performance is most likely to fail to carry over to new slides.

CC on AI Liability and RecourseMechanisms exist that let someone challenge an output, such as audit trails, source traceability or review before commit, with nothing standing behind the output and no route for the harmed party.
Vendor Published

Users see each model's validation figures before deploying it, and HistAI's liability is capped. The agent reports AUROC, sensitivity, specificity and a confusion matrix on a held out validation split, and every job keeps an audit trail of its configuration, cohort and metrics.

HistAI's terms (last updated 10 April 2026) cap its liability at $1,000 for any cause and keep custom models as HistAI's property. No warranty, error commitment or remedy attaches to model outputs.

Integration and Deployment
DD on EHR and Interoperability DepthNo integration evidence. A connector described as available on request grades here until one exists.
Vendor Published

There is no published integration with laboratory or clinical systems. Custom models deploy to HistAI's own CellDX platform, where slides are uploaded and viewed, and an API key lets an AI agent drive training and deployment. Business plan subscribers can download weights to run elsewhere. No laboratory information system, image management system, scanner, standard or interface is named for bringing slides in or sending results out.

BB on Deployment Model and Data ResidencyOptions and residency are stated with isolation or the processing path left open.
Vendor Published

Two deployment routes are described, and the hosted route's region is not stated. Models train and run on HistAI's CellDX cloud, which the launch post describes as secured on Microsoft Azure with managed identities and private storage endpoints. Business plan subscribers can download trained weights and run them in their own environment, so slides analyzed there need not reach HistAI.

No Azure region, data location or separation between customers on the hosted platform is published.

Commercial
AA on Commercial TransparencyPublished tiers with figures, a stated unit of charge, and a route to start without a sales conversation.
Vendor Published

Prices, units and a self service route are all published. HistAI lists subscription plans from $15 to $39 a month, with a $20 credit on registration, and model training from $15 to $50 a model, charged for the GPU time actually used; its launch post puts GPU time at $8 an hour and a typical model at about $5 to $10. A feature store of more than 66,000 slides is free for training, extra runs are sold as prepaid Run Packs, and sign up and payment happen online. Data Hub slides are priced separately, per slide.

CC on Setting and Specialty CoverageCoverage is claimed broadly without specifics, or stated clearly with nothing validating it yet.
Vendor Published

No limit is set on where AI Autopilot applies, and nothing validates its use. The training feature store holds more than 32,000 cases and 66,000 H&E slides across breast, colorectal, gastrointestinal and other tissues, and users can run their own slides through deployed models. It is offered to pathologists, researchers, data scientists and laboratories.

HistAI does not publish where custom models built this way perform well or poorly, and its terms leave validation for any clinical use to the laboratory.

Alternatives

Alternatives to HistAI AI Autopilot

Other products that do the same kind of work as HistAI AI Autopilot, and what each one does.

All HistAI AI Autopilot alternatives, with prices →

Bioptimus is a Paris company that builds AI foundation models for pathology and biology. Drug companies, research groups and pathology software makers license them as a base for their own tools.

Gestalt sells PathFlow, a digital pathology platform that private laboratories, health networks and academic centers use to bring slides, case and patient information and AI results into one workspace.

Imagene AI reads routine stained lung cancer slides and predicts which gene changes a tumor carries, without using more tissue for sequencing.

Aiosyn sells AI that analyzes images of stained tissue, to pathology laboratories and to drug companies running animal studies.

Biomy makes DeepPathFinder, cloud software that uses AI to map immune and cancer cells on tissue slides for drug research.

Techcyte sells Fusion, a digital pathology platform that human, veterinary and environmental laboratories use for both anatomic and clinical pathology.

Commercial

Pricing

Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.

Entry Price Pricing Basis BAA Tier Implementation Source
From $15 a month; model training from $15 a model
$15 baseline
Self service subscription plans from $15 to $39 a month; custom model training from $15 to $50 a model on the product page, charged for GPU time at $8 an hour; free training access to a feature store of more than 66,000 slides; extra model runs sold as prepaid Run Packs. No business associate agreement published; the privacy policy claims HIPAA and GDPR compliance and the terms make the user responsible for compliant uploads. None published; sign up is self service. Vendor Published

HistAI publishes subscription tiers and a per model training price tied to GPU time, so a buyer can estimate cost before contact. Its May 2026 launch post gives GPU time at $8 an hour and a typical model at about $5 to $10, below the $15 to $50 on the product page. Data Hub slides are licensed separately per slide, from $2 to $5 for H&E and $23 to $40 for IHC depending on volume. Custom models remain HistAI's property under its terms.

HistAI AI Autopilot pricing in full →