Lasso Security Pricing: Quote Scope and Open-Source Boundaries

Lasso Security pricing starts with a sales conversation. Lasso's public platform page describes an AI security lifecycle spanning discovery and assessment, followed by runtime protection, while its commercial route asks the buyer to book a live demo. It publishes no complete dollar rate card for that platform. The separate open-source MCP Gateway changes the scope discussion because source availability for one gateway does not make the commercial platform free.

TL;DR

• Lasso publishes platform scope and a live-demo route, rather than a complete public price list.
• The commercial platform covers discovery and risk assessment, with red teaming plus runtime enforcement and detection activities.
• Lasso's MCP Gateway is open source and available through GitHub, but it is one product boundary rather than the entire platform.
• Ask for a written quote that separates modules and deployment points; protected applications and support; retention and any usage limits.

Lasso Security pricing is quote-led

The Lasso AI Security Platform page presents a commercial platform and sends buyers to a demo. The live demo page asks for buyer details and promises a personalized walkthrough. Neither page displays a complete public dollar schedule, a per-user price, or a standard enterprise contract value.

That leaves procurement with a scoping job. A buyer should identify which parts of Lasso's lifecycle enter the proposed deployment before asking for a number. Discovery and inventory create a different workload from adversarial testing. Runtime inspection creates another, especially when traffic crosses several applications and model providers, including MCP servers.

I would reject a Lasso quote labeled only "AI Security Platform." That line hides the exact modules and operating assumptions finance will need at renewal.

Platform modules define the commercial scope

Lasso groups its current platform around discovering agents and applications. It also covers assessing risk and protecting activity at runtime. The page names AI security posture management and automated AI red teaming, plus runtime enforcement and AI detection and response. Those functions can touch separate stages of an application's lifecycle and separate technical owners.

Build the request around the work each module performs. Inventory scope should name the cloud platforms and agent builders; applications and tools; connected MCP servers used for discovery. Assessment scope should identify the applications and release environments examined. Runtime scope should specify the proxies and APIs where policy will run, including AI gateway paths.

A diagram helps. Draw separate rectangles for the application and model endpoint. Add another rectangle for every MCP server. Put a red dot on each place Lasso will inspect or enforce. The resulting page gives the vendor enough detail to return a quote that an architect can verify.

The open-source MCP Gateway is a separate boundary

Lasso announced its open-source MCP Gateway on 15 April 2025 and linked the project on GitHub. The announcement describes a proxy and orchestrator for MCP interactions, with policy and monitoring that includes attack blocking. It also says the gateway is available as part of Lasso's platform.

The gateway code has an open-source route. The wider Lasso platform remains a commercial product reached through a demo. Procurement should ask which controls run in the open-source gateway and which require the commercial platform. It should also ask which services come with an enterprise agreement.

The MCP security tools guide separates gateway controls from scanners and registries. MCP traffic control cannot substitute for application discovery, red-team automation, or a broader detection program merely because the products share a vendor name.

Quote inputs should follow the deployment

A useful quote request names the protected estate and the operating obligations. Start with the number and type of AI applications and agents, then MCP clients and MCP servers. Add the deployment environments and model providers. State where runtime enforcement will sit and who will operate that path.

Then ask Lasso to itemize the commercial response:

• Included platform modules and any module-specific capacity.
• Application and agent limits; connector and user limits; request or environment limits used for billing.
• Hosted and customer-managed components, including required infrastructure.
• Data retention and evidence export; onboarding and support; response terms.
• Red-team cadence across test environments and treatment of usage above contracted limits.
• Contract term and renewal basis, plus the definition of every billable unit.

Avoid converting those inputs into an invented per-request price. The vendor's written quote should supply the rate and explain which unit it applies to.

Content security and authorization records remain distinct

Lasso describes broad AI security coverage, including discovery and runtime threat prevention. A regulated model route may also need a decision record binding an authenticated caller to a data classification and destination, with the policy version and outcome. The two requirements can overlap at an HTTP proxy while serving different reviews.

The DeepInspect and Lasso comparison examines that architectural split. For a pricing exercise, keep the records separate. Price Lasso against the modules and assets it will cover. Price an identity-bound policy layer against the model requests and evidence obligations placed on that layer.

Combining the products under one "AI security" budget line makes later attribution harder. A security leader should be able to point to the invoice and identify which amount pays for discovery and which amount pays for runtime detection. The invoice should also identify which control produces the authorization record used during an audit.

DeepInspect

DeepInspect governs authenticated HTTP traffic deliberately routed between users or agents and LLM endpoints. It evaluates the application-supplied identity and role, plus data classification and destination, against versioned policy, then records the decision before forwarding allowed traffic.

That boundary can sit beside Lasso's broader discovery and testing, plus runtime security functions. Scope each control against the path it can inspect and the evidence it must produce. Book a technical deep dive at deepinspect.ai.