Glass AI (clinical decision support, differential diagnosis & ambient scribing)
Glass Health · Ambulatory and hospital point-of-care, telehealth, clinician workflow · SaaS (freemium), enterprise instances, developer API
Clinical Evidence & Decision Support
Glass AI is a clinician-facing generative AI platform that produces ranked differential diagnoses and evidence-based clinical plans from a patient summary, with every answer citing the sources it reasoned from. It also provides ambient scribing and documentation support and is designed for clinician supervision at the point of care, with health-system instances and a developer API.
Evidence at a glance
Freshness, source-checked counts, confidence and gate results below reflect the frozen evidence snapshot through . A later editorial review does not update these results.
- Stated intended users
- Physicians, NPs, PAs, residents, and medical students; health systems; digital health platforms via developer API
- Documentary gate at snapshot
- Failed
- Scenario match
- No evaluated buying scenario currently covers this product
- Comparability
- Not yet comparable in this scenario
- Decision readiness at snapshot
- HISTORICAL — Research only
- Documented evidence at snapshot
- 0 of 4 evidence claims source-checked · confidence Low
- Critical unknowns
- scaled production deployment · security and governance evidence · documented economic outcomes
- Latest checked-source retrieval in snapshot
- not yet measured
Expand each claim to inspect its source, dates and recorded verification status.
vendor pageIn the physician-validated Harvard-Stanford ARISE NOHARM clinical safety benchmark, Glass ranked among the top medical AI systems, ahead of …2026-08-15
Exact claim. In the physician-validated Harvard-Stanford ARISE NOHARM clinical safety benchmark, Glass ranked among the top medical AI systems, ahead of models from OpenAI and Anthropic; Glass reasons over 38M+ peer-reviewed citations, current clinical guidelines, and 150,000+ FDA drug labels with cited answers
- Source
- glass.health
- Type
- vendor page
- Published
- 2026-08-15
- Retrieved
- 2026-08-27
- Verification
- unverified
- Confidence
- not stated
- Reviewer
- healthit-gate-v1
- Claim version
- v1
- Supports
- evidence_strength
- Score contribution
- Feeds evidence strength
Self-reported by the vendor. Independent verification is pending unless marked otherwise.
Open glass.healthvendor pageThe July 2025 announcement introduced the wide release of DeepConsult, with the feature free to verified U.S. clinicians, expanding Glass AI…2025-07-31
Exact claim. The July 2025 announcement introduced the wide release of DeepConsult, with the feature free to verified U.S. clinicians, expanding Glass AI's clinical reasoning capabilities
- Source
- glass.health
- Type
- vendor page
- Published
- 2025-07-31
- Retrieved
- 2026-08-27
- Verification
- unverified
- Confidence
- not stated
- Reviewer
- healthit-gate-v1
- Claim version
- v1
- Supports
- deployment_traction
Self-reported by the vendor. Independent verification is pending unless marked otherwise.
Open glass.healthpress releaseGlass AI launched a Developer API for its clinical decision support platform, enabling third parties to build custom integrations (including…2025-02-28
Exact claim. Glass AI launched a Developer API for its clinical decision support platform, enabling third parties to build custom integrations (including ambient scribing workflows)
- Source
- linkedin.com
- Type
- press release
- Published
- 2025-02-28
- Retrieved
- 2026-08-27
- Verification
- unverified
- Confidence
- not stated
- Reviewer
- healthit-gate-v1
- Claim version
- v1
- Supports
- interoperability
- Score contribution
- Feeds interoperability
Vendor or press announcement. Verify material claims against primary filings or contracts.
Open linkedin.comindependent reportGlass Health (YC W23) launched an LLM-powered tool that generates differential diagnoses and evidence-based treatment options from clinician…2023-09-08
Exact claim. Glass Health (YC W23) launched an LLM-powered tool that generates differential diagnoses and evidence-based treatment options from clinician-entered patient summaries, with roughly $6.5M in total funding including a $5M seed led by Initialized Capital
- Source
- techcrunch.com
- Type
- independent report
- Published
- 2023-09-08
- Retrieved
- 2026-08-27
- Verification
- unverified
- Confidence
- not stated
- Reviewer
- healthit-gate-v1
- Claim version
- v1
- Supports
- general
Independent source. Methodology and sample vary by report; read the underlying method.
Open techcrunch.comVerified provider experience is not yet publishable. Fewer than five independent provider organizations have contributed eligible reviews. Reviewer mode is off. Experience never changes documentary scoring, freshness, or rank.
Buyer lenses — evidence-backed bands at snapshot (not a 0–100 rank)
Deployment, integration, data handling at snapshot
- Integration maturity
- announced — Announced or documented integration; marketplace listing not confirmed.
- EHR path
- UNKNOWN
Why unknown
Why. No reliable source on file. The platform does not infer values from weak sources.
What would close it. A primary source or independent verification.
- FHIR / API
- Documented
- SSO
- UNKNOWN
Why unknown
Why. Single-sign-on support is not documented in vendor or marketplace materials on file.
What would close it. Vendor security documentation or marketplace listing mentioning SSO/SAML/OIDC.
- PHI / BAA
- Clinician-entered patient data processed under enterprise data agreements; clinician-supervised outputs
- Security attestations
- UNKNOWN
- FDA status
- UNKNOWN
Why unknown
Why. No FDA listing found for this product's functions in the tracked category.
What would close it. An FDA 510(k)/De Novo listing, or a clear statement that no regulated function is claimed.
- Pricing
- Not publicly listed
Technical provenance
snap_36bba25c577cbb44 · Evidence cutoff: 2026-08-27T04:05Z · Snapshot generated: 2026-10-21T19:39Z · build ffbf5c3
Forecast calibration is not in the public snapshot — shown as UNKNOWN rather than read from SQLite.
Known limitations
- Deployment traction is UNKNOWN.
Why unknown
Why. No reliable source on file. The platform does not infer values from weak sources.
What would close it. A primary source or independent verification.
- Trust & governance is UNKNOWN.
Why unknown
Why. No reliable source on file. The platform does not infer values from weak sources.
What would close it. A primary source or independent verification.
- Economic value evidence is UNKNOWN.
Why unknown
Why. No independently verifiable ROI claim (peer-reviewed, audited case, or independent benchmark) is on file.
What would close it. An audited implementation report or peer-reviewed study with payback/throughput numbers.