GLACIS
glacis.io
- Free plan
- Web
- API
- Self-hosted
What GLACIS is
GLACIS places runtime controls in the path of AI actions and creates signed records of covered decisions for external verification. It can allow actions through unchanged or, when they cross configured bounds, constrain them, hold them for a person, or deny them. Glacis says it operates nine production components as a managed service while working with a customer’s existing identity, cloud, and security systems. Decisions produce signed records in OVERT, a published open format that implementers and reviewers can inspect and use without a proprietary schema. A verifier can check a record in a browser without a Glacis account. The site says payloads stay in the customer’s environment and records carry hashes; whether protected fields can be omitted depends on deployment setup. A record check does not establish whether an action was correct, every event was captured, or a system is safe or compliant. GLACIS offers a free account at no cost, with no card required, and describes a six-week evaluation on the workflow being reviewed. It is available via API, self-hosting, and web.
Who it suits
GLACIS is aimed at AI sellers serving regulated buyers, healthcare AI overseers, and AI-agent builders or operators. Its signup form also includes people evaluating the product.
Good to know, and what to watch
- Can hold or deny actions outside configured bounds.
- Signed records use the published OVERT format.
- Records can be checked in a browser without an account.
- Free account requires no card.
- No automated evidence-pack export is offered today.
- Evidence packs are assembled by hand for a review.
- Record checks do not establish correctness or compliance.
Plans
Every plan and priceFree account
Free plan · 14-day trial
Getting started
- Open the GLACIS website
- Create a free account at no cost; no card is required
- Choose API, web, or self-hosted deployment
- Connect the identity provider, cloud, and security systems already in use
- Arrange a six-week evaluation for the workflow under review
Limits to check
The free offering includes a 14-day trial. Evidence packs are not exported automatically and must be assembled by hand; record verification does not establish action correctness, complete capture, safety, or compliance.
The Club's review
GLACIS is a runtime control layer for teams building, selling, or overseeing AI workflows, particularly where regulated customers or consequential agent actions are involved. It is most compelling when intervention must happen before an action proceeds and when an independently checkable record matters; teams needing automated review packages will need extra process.
Overview
GLACIS sits in the path of AI actions and evaluates them against configured bounds. An action can pass unchanged, be constrained, be held for a person, or be denied. That makes it more than a retrospective logging tool: operators can introduce a decision point before a risky operation continues.
The accompanying record is signed, but its scope is important. Verification confirms the record’s integrity; it does not show that the action was correct, that every event was captured, or that the overall system is safe or compliant. Buyers should treat GLACIS as a control and evidence component, not a complete assurance program.
Key features
Runtime protection and deployment
Prompt-injection defense, PII redaction, and output guardrails address several practical failure modes in AI applications. The hybrid deployment model supports API, web, and self-hosted use. GLACIS says its managed service operates nine production components while integrating with the identity provider, cloud, and security systems already in place. That suits organizations wanting managed operations without replacing core infrastructure, although the value depends on how well the configured bounds reflect each workflow.
Signed records and OVERT
Records use OVERT, a published open format that implementers and reviewers can inspect and use without adopting a proprietary schema. A verifier can check a record in a browser without a GLACIS account, which lowers friction for an external reviewer. Payloads remain in the customer environment, while records carry hashes; protected fields can be omitted depending on deployment configuration. This design can reduce unnecessary data exposure, but it also makes deployment choices part of the evidence model.
Evidence workflow
GLACIS does not currently export evidence packs automatically. Packs must be assembled manually for a particular review, so teams with recurring audits or many workflows should budget analyst time and define a repeatable collection process. The company describes a six-week evaluation on the workflow being reviewed, a useful way to judge fit before committing to a managed deployment.
Pricing
GLACIS uses a freemium model. The Free account costs free, requires no card, and takes about a minute to set up. A 14-day trial is also offered. The free account is the sensible starting point for evaluators and small teams validating controls or record verification; the published terms do not provide paid-plan prices or quotas, so organizations planning broader managed use should request a commercial proposal.
The free entry point lowers the cost of an initial workflow review, but it does not remove the operational trade-offs: evidence packs remain manual, and verification has the limits described above. Teams comparing plans should confirm seats, usage allowances, renewal terms, and deployment scope before treating a trial result as a production commitment.
Platforms
GLACIS supports API, web, and self-hosted deployment. Its hybrid model is suited to teams that need customer-environment data handling alongside a managed control plane.
Who it's for
The strongest fit is an AI vendor serving healthcare or other regulated buyers, a healthcare AI overseer, or a builder or operator of AI agents whose actions need runtime intervention. It also fits reviewers who need an external party to verify signed records without creating a GLACIS account. It is less suitable for organizations seeking turnkey, automatically generated evidence packages or a product that alone proves correctness, full event capture, safety, or compliance.
Pros and cons
Pros
- Runtime choices before execution: actions can be allowed, constrained, paused for human review, or denied at configured boundaries.
- Independently checkable records: browser verification without an account helps external reviewers inspect covered decisions.
- Open record format: OVERT can be inspected and used without dependence on a proprietary schema.
- Customer-environment data handling: payloads stay in the customer environment and records use hashes, with optional omission of protected fields.
- Managed integration: nine production components are operated as a service while working with existing identity, cloud, and security systems.
Cons
- Manual evidence packs: each review package must be assembled by hand, limiting automation for repeat audits.
- Verification is narrow: a valid record does not establish correctness, complete capture, safety, or compliance.
- Commercial scope is unclear: published information gives a free account and trial but no paid-plan pricing or quotas.
- Selective managed capacity: GLACIS says it takes on a small number of managed deployments each quarter, which may constrain timing.
- No bug bounty: the vulnerability policy describes response targets but does not offer bounty payments.
Alternatives
For a broader shortlist, see LLM Security Tools.
- Prompt Inspector is a better fit when you want a free tier with 1,000 detections per month plus REST API, SDK, and MCP access.
- NVIDIA NeMo Guardrails suits teams wanting a free, open-source Python package and willing to set up external models and services.
- Amazon Bedrock Guardrails is for buyers who prefer a paid AWS service with metered text, image, and denied-topic filters.
- PromptGuard is preferable when a free plan with 20,000 scans a month, one API key, one project, and 24-hour log retention is the priority.
- GuardionAI is an enterprise paid option with a free trial and custom contract terms.
- LLM Guard is a free open-source option, but its repository is archived and no longer actively maintained.
- Openlayer Guardrails fits teams seeking a free workspace with project, pipeline, inference-log, test, and retention limits.
- VotalAI is another paid platform that offers a free plan and free trial.
Verdict
Choose GLACIS if you need runtime intervention for AI actions and a signed, open-format record that outside reviewers can verify without an account. Look elsewhere if automated evidence-pack generation or stronger assurances about correctness and complete event capture are requirements.
The verdict
GLACIS adds controls to AI actions and creates signed records that can be checked independently. Its verification limits and manual evidence-pack process matter for teams evaluating it for review workflows.
Questions people ask
Does GLACIS have a free plan?
Yes. The free account costs nothing, requires no card, and has setup of about a minute. The pricing note also describes a 14-day trial.
Which deployment options are available?
GLACIS is listed for API, self-hosted, and web use, with a hybrid deployment model.
What is OVERT?
OVERT is a published open record format used for signed decision records. Implementers and reviewers can inspect and use it without a proprietary schema.
Can someone verify a record without GLACIS?
Yes. A verifier can check a record in a browser without a GLACIS account.
Does GLACIS export evidence packs automatically?
No. Evidence packs are assembled manually for a particular review.
What does GLACIS do with payloads?
The site says payloads stay in the customer’s environment and records carry hashes. Whether protected fields can be left out depends on deployment setup.
What the maker publishes
- Control decisions
- For actions that cross configured bounds, Glacis can constrain, hold for a person, or deny the action; it can also allow actions through unchanged. source
- Data handling
- The site says payloads stay in the customer’s environment and records carry hashes; protected fields can be left out, depending on deployment setup. source
- Deployment evaluation
- The contact page describes a six-week evaluation on the workflow a customer is reviewing. source
- Evidence pack limitation
- Glacis does not offer automated evidence-pack export today; packs are assembled by hand for a particular review. source
- Managed service
- Glacis says it runs nine production components as a managed service, while working with the identity provider, cloud, and security systems the customer already uses. source
- Open standard
- OVERT is a published open record format that implementers and reviewers can inspect and use without a proprietary schema. source
- Records
- Records are signed and use OVERT, an open record format; a verifier can check a record in a browser without a Glacis account. source
- Security reporting
- The vulnerability disclosure policy says Glacis aims to acknowledge reports within three business days and provide an initial assessment within ten business days; it offers no bug bounty. source
- Support
- Glacis says it typically replies to contact requests within one business day and takes on a small number of managed deployments each quarter. source
- Verification limits
- A record check does not establish whether an action was correct, every event was captured, or a system is safe or compliant. source
- What it does
- Glacis places runtime controls in the path of AI actions and creates signed records of covered decisions for external verification. source
- Who it serves
- The signup form lists AI sellers serving healthcare or other regulated buyers, healthcare AI overseers, AI-agent builders or operators, and people evaluating the product. source
Checked against the maker's pages on 4 Oct 2026.