Prisma AIRS AI Gateway
Prisma AIRS AI Gateway centralizes control, security, and visibility for requests that applications and teams send to large language models.
What Prisma AIRS AI Gateway is
Prisma AIRS AI Gateway centralizes control, security, and visibility for requests that applications and teams send to large language models. It connects with more than 1,600 models across at least 50 providers, including major hosted and self-hosted options. Developers use the standard OpenAI API format, allowing provider changes and fallback models without application code changes. Workspaces manage model access, budgets, rate limits, and guardrail policies through scoped virtual keys. Inline controls can block prompt injection, detect sensitive data, and restrict topics in requests and responses. Integrations cover LLM providers, MCP servers, secret references, and agent backends. Credentials are encrypted at rest and decrypted only in memory. Deployment is available as Palo Alto Networks SaaS or Hybrid on a customer-managed Kubernetes cluster through Strata Cloud Manager; Hybrid keeps payloads in the customer environment and requires Kubernetes and Helm 3. Pricing is on request and based on token usage.
Good
- Connects to over 1,600 LLMs
- Standard OpenAI API compatibility
- Budget, rate, and access controls
- Prompt-injection and sensitive-data protections
- Hybrid deployment keeps payloads customer-side
Watch out
- Pricing requires contacting sales
- Hybrid deployment requires Kubernetes and Helm 3
- Paid service
Who it suits
It suits teams governing LLM access, spending, security controls, and observability across multiple providers or deployments.
Verdict
Prisma AIRS AI Gateway provides centralized model access, governance, runtime protections, and analytics. Organizations can choose hosted SaaS or Hybrid deployment, with token-based licensing and sales-managed credit requirements.
Plans and prices
From Prisma AIRS AI Gateway's own pricing page, checked 2 Oct 2026.
| Plan | Price | What it gives you |
|---|---|---|
| Consumption-based AI Gateway licensing | Priced on request | Token metering · Tiered volume discounts · Applies to LLM, MCP, and A2A traffic · Software NGFW credits based on token usage; contact your sales representative for the credits required. |
The facts
Each one read from Prisma AIRS AI Gateway's own pages, with the date it was checked.
What it does
- API compatibility
- Developers send requests using the standard OpenAI API format, and provider switches or fallback models do not require application code changes.checked 2 Oct 2026
- Credential security
- Integration credentials are stored encrypted at rest and decrypted only in memory during requests, while raw API keys remain hidden from workspaces.checked 2 Oct 2026
- Deployment
- The gateway offers SaaS deployment hosted by Palo Alto Networks and Hybrid deployment on a customer-managed Kubernetes cluster; both are managed through Strata Cloud Manager.checked 2 Oct 2026
- Governance
- Workspaces let teams manage LLM access, budgets, rate limits, and guardrail policies, with scoped virtual keys controlling model access and spending.checked 2 Oct 2026
- Hybrid data handling
- In Hybrid mode, LLM request and response payloads are processed within the customer's environment and are not transmitted to Palo Alto Networks.checked 2 Oct 2026
- Hybrid requirements
- Hybrid deployment requires a provisioned Kubernetes cluster and Helm 3.checked 2 Oct 2026
- Identity and encryption
- Admin settings support SSO with custom OIDC providers and integration with an organization's KMS for managing secrets under its own key hierarchy.checked 2 Oct 2026
- Integrations
- The integration management surface covers LLM providers, MCP servers, secret references, and agent backends; listed LLM options include OpenAI, Anthropic, AWS Bedrock, Azure OpenAI, Google Gemini, and self-hosted models.checked 2 Oct 2026
- Pricing limits
- The maker does not publish an exact price on its licensing page; it says credit requirements depend on anticipated token usage and are calculated by the sales team, with tiered volume discounts.checked 2 Oct 2026
- Provider support
- It connects to more than 1,600 LLMs across 50 or more providers, including OpenAI, Anthropic, Google, AWS Bedrock, Azure OpenAI, Mistral, Cohere, and self-hosted models.checked 2 Oct 2026
- Purpose
- Prisma AIRS AI Gateway centralizes control, security, and observability for AI requests made by teams and applications to LLMs.checked 2 Oct 2026
- Runtime protection
- Inline guardrails can block prompt injection, detect sensitive data, and enforce topic restrictions on requests and responses.checked 2 Oct 2026
The company
- Company
- Palo Alto Networks headquarters are at 3000 Tannery Way, Santa Clara, California, and the company was incorporated in 2005.checked 2 Oct 2026
- Founded
- 2005checked 28 Sep 2026
- Headquarters
- Santa Clara, California, United Stateschecked 28 Sep 2026
Alternatives to Prisma AIRS AI Gateway
- 1
LangSmith
Free planWebAPILinuxmacOS8.0$39 usage-based - 2
Manifest
Free planOpen sourceWebAPILinuxSelf-hosted7.9From $19 a month - 3
Bifrost
Free planOpen sourceWebAPILinuxmacOS7.6Free plan - 4
TrueFoundry AI Gateway
Free planWebAPILinuxSelf-hosted7.6From $499 a month - 5
YoloRouter
Free planOpen sourceWebAPILinuxmacOS7.5Free plan - 6
Eden AI
WebAPI7.4$5.50 one-time