How much does Latent Health cost?

Enterprise sales; no public rates.

What are the best alternatives to Latent Health?

Based on documented capability coverage across the Agentic Index 14 point taxonomy, the closest alternatives to Latent Health among healthcare agents are Humata Health, AlethianAI, Claim Health, CombineHealth. Each is graded on the same public evidence, and no vendor pays for placement.

Does Latent Health give you tracing and observability into agent runs?

Full under the Agentic Index verification standard. Latent Health publishes public evidence of tracing, observability and audit logging into agent runs, so a team can see what an agent did and why. Evidence basis: Every output traces back to the chart with a built-in audit trail recording each step of a case (chart note read, criteria matched, prior auth prepared, pharmacist signed), with queue management and outcomes, compliance and financial dashbo. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Latent Health require human approval before an agent acts?

Full under the Agentic Index verification standard. Latent Health publishes public evidence of human approval steps and guardrails, so an agent action can be held for review before it executes. Evidence basis: A pharmacist decides each case after Latent prepares and fills it, choosing to approve, send back or ask the prescriber. The pharmacist's signature is a recorded step before submission. Latent prepares, fills and submits the case, and pharm. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Latent Health have SOC 2, SSO and role based access control?

Partial under the Agentic Index verification standard. Latent Health publishes some public evidence of security and identity controls, but narrower than the full set of SOC 2, single sign on (SSO) and role based access control (RBAC). Evidence basis: Latent claims SOC 2 and HIPAA compliance and links a Vanta trust center. It names no SSO, roles or permission model for the customer's users. CIPHER can remove patient identifiers from notes as they are written rather than in an overnight b. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Back to vendors

Latent Health

Q: Who is Latent Health for?

health systems, pharmacies, providers

Visit site

Healthcare agentindependentVerified 2026-09-29

Entry priceContact salesFull pricing detail

Enterprise pharmacy intelligence for health systems. A clinical agent engine reads charts, claims and payer policy to find eligible patients, prepare and submit prior authorizations and appeals for pharmacist sign-off, audit eligibility and check in with patients, at 60+ health systems.

Latent is an enterprise pharmacy intelligence platform used by more than 60 US health systems, including Mount Sinai, Cleveland Clinic, Yale New Haven, Mayo Clinic and Ochsner, across more than five million patient medication cases a year. One clinical agent engine, built on in-house LLMs and retrieval, reads the chart, claims, payer policies and drug labels and runs each case through four stages.

They are Identify (population scans that surface patients eligible for therapy), Access (prior authorization, clinical review, appeals, benefit routing and payer outreach), Treat (eligibility compliance and leakage analytics for discount and assistance programs) and Engage (routine patient check-ins, with pharmacists pulled in when a patient flags a problem). Latent prepares and fills each case and a pharmacist approves, sends it back or asks the prescriber, and every output traces back to the chart in a built-in audit trail. It is chart-native and EHR-agnostic, states SOC 2 and HIPAA compliance, and has raised $80 million from Transformation Capital, Spark Capital, Conviction, General Catalyst and Y Combinator. Pricing is by sales conversation.

Vendor details

Canonical URL

https://www.latenthealth.com

Funding status

$600M valuation (Forbes, Mar 2026)

Company status

independent

Use cases & customers

Primary use cases

prior authorization evidence compilationclinical record question answeringpayer status callsdrug access workflows

Target customers

health systemspharmaciesproviders

Deployment options

SaaS

Integrations

Latent is chart-native and EHR-agnostic. It works across the EHR, pharmacy system, payer portals and claims, submitting prior authorizations to payers, and names no specific connectors.

In practice

Your specialty pharmacy team spends hours on prior authorizations. Latent reads the chart and the payer's policy, fills and submits the form, and leaves the decision and signature to your pharmacist.

Patients eligible for a specialty therapy go unnoticed across your health system. Latent scans charts and claims to surface them and starts the access work.

A payer denies a prior authorization. Latent drafts the appeal with evidence from the chart, and your pharmacist decides whether it goes out.

Sources & related URLs

Research sources

https://www.forbes.com/sites/rashishrivastava/2026/03/18/this-600-million-ai-startup-is-speeding-up-paperwork-so-patients-get-drugs-faster/

Agentic Index coverage score

7.5 / 14 capabilities · 54%

Integrations & Tool Calling	Full
Latent works in the chart and with any EHR, across the EHR, the pharmacy system, payer portals and claims. It fills and submits prior authorization forms to payers and runs payer outreach. Latent names no specific EHR connectors. It goes live in a day. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Workflow Orchestration	Full
One clinical agent engine runs each case through Identify, Access, Treat and Engage. Its branches include site of care and benefit routing, appeals after denial, a pharmacist's decision to approve, send back or ask the prescriber, and patient check-ins routed to a pharmacist when an adverse effect is flagged. The engine reads, reasons and acts across pharmacy workflows. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Knowledge Grounding & RAG	Full
The clinical agent engine reads charts, claims, payer policies and drug labels, parses the payer's policy, checks medical necessity against it and finds the supporting evidence, with every output tracing back to the chart. Clinical search runs on in-house retrieval. CIPHER, Latent's redaction model, removes the 18 HIPAA identifier types from a clinical note while keeping its medical meaning. SourceLatent Health, latenthealth.com/platform and /cipherread 2026-10-05
Human Oversight & Guardrails	Full
A pharmacist decides each case after Latent prepares and fills it, choosing to approve, send back or ask the prescriber. The pharmacist's signature is a recorded step before submission. Latent prepares, fills and submits the case, and pharmacists keep the decisions that belong to them. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Security, Identity & Governance	Partial
Latent claims SOC 2 and HIPAA compliance and links a Vanta trust center. It names no SSO, roles or permission model for the customer's users. CIPHER can remove patient identifiers from notes as they are written rather than in an overnight batch. SourceLatent Health, latenthealth.com/platform and /cipherread 2026-10-05
Observability & Auditability	Full
Every output traces back to the chart with a built-in audit trail recording each step of a case (chart note read, criteria matched, prior auth prepared, pharmacist signed), with queue management and outcomes, compliance and financial dashboards. Sourcelatenthealth.com/platformread 2026-09-29
Memory & State Persistence	Not documented
The agents have no memory that Latent describes, and it does not say what would be kept or for how long. A case carries its own chart evidence through the four stages. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Deployment & Data Residency	Not documented
No hosting region is published, and Latent names no region choice, customer environment or on premises option. It goes live in a day inside the EHR. Its models are served on Baseten with sub second latency and 99.999% availability. SourceLatent Health, latenthealth.com/platform and the Baseten case study press pageread 2026-10-05
Prebuilt Agents, Templates & Packs	Partial
Four modules (Identify, Access, Treat, Engage) carry named capabilities such as care gap identification, prior authorization, appeals and eligibility compliance. They are one engine running the same case through four stages, so they work as stages of one workflow, not separate products. Latent publishes no templates or agent library. Treat audits eligibility and reduces leakage, and Engage runs routine patient check ins with pharmacist escalation. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Triggers & Channel Coverage	Full
Population scans read charts and claims to surface eligible patients without a person starting each case, and the engine runs routine patient check-ins and payer outreach on its own. More than 60 major US health systems run the platform. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Model Flexibility & Routing	Not documented
Models are Latent's own. Its LLMs and retrieval are built in house and served on Baseten, and it also uses distilled models such as CIPHER, a 1.5B parameter token classifier trained on labels from an ensemble of eight open source LLMs. Latent names no customer model choice or routing. SourceLatent Health, latenthealth.com/cipher and the Baseten case study press page; latenthealth.com/press/baseten-case-studyread 2026-10-05
APIs, SDKs & MCP Extensibility	Not documented
No API, SDK or MCP server is published for customers to build on. Latent works inside the customer's EHR, pharmacy system, payer portals and claims instead. SourceLatent Health, latenthealth.com/platformread 2026-10-05
Testing, Debugging & Optimization	Not documented
CIPHER's accuracy comes from Latent's own model benchmark, with 97.9% recall and 93.85% F1 on 195 expert annotated notes, beating its teacher model on recall. Latent names no evaluation harness, test set or quality gate the customer runs. It reports 80% less time to draft appeals and 45% more volume without added headcount. SourceLatent Health, latenthealth.com/cipher and /platformread 2026-10-05
Browser & Computer Use	Partial
Payer portals are among the systems Latent works in when it fills and submits prior authorizations, so it acts inside platforms that have no API. Latent does not say how it operates the portals. The same engine also handles payer communications and appeals. SourceLatent Health, latenthealth.com/platformread 2026-10-05

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Product updates

September 15, 2026·Agent capabilityVerified

Latent Health introduced autonomous clinical agents that speed patient access to specialty therapies such as GLP-1 medications. The agents evaluate eligibility and gather the clinical information needed for approval.

Bears on: Agent capability

View source

View all Latent Health updates →Tracked since Sep 2026 · Verified from public vendor sources

Pricing

Contact sales

What is public

No prices, plans or billing basis are published; contact is the route. Public figures are outcomes from health system deployments.

Variable cost rationale

No billing basis or rate is published, so exposure to case volume cannot be checked from public pages.

Sales call required

Yes, required for paid access

Free / trial

None public

Agentic Index verified 2026-09-29

Official pricing page

Related vendors

Alternatives to Latent Health

The closest documented capability profiles to Latent Health among healthcare agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

Humata Health7.5 / 14Adds documented Testing, Debugging & Optimization
AlethianAI7.0 / 14Fuller documented coverage on Prebuilt Agents, Templates & Packs
Claim Health6.0 / 14A lighter documented profile than Latent Health
CombineHealth9.0 / 14Adds documented Testing, Debugging & Optimization
Alaffia Health8.5 / 14Adds documented Deployment & Data Residency
Anterior8.5 / 14Adds documented APIs, SDKs & MCP Extensibility and Testing, Debugging & Optimization

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Compare Latent Health with other vendors →Compare all healthcare agents →See where Latent Health ranks →

Data confidence: medium

Head to head

Latent Health vs Mandolin