AI Tools
PRISMA AIRS AI Gateway
PRISMA AIRS AI Gateway: precios, capacidades, actualizaciones y alternativas, leídos en las páginas del propio desarrollador.

En inglés
Formerly Portkey AI Gateway
Key facts
- Price
- Not stated on the maker’s pages
- Free plan
- Yes Read on portkey.ai
- Platforms
- Web · Linux · API
- API
- Yes
- Open source
- Yes Read on github.com
- Last update
- 29 Apr 2026 Read on new.portkey.ai
PRISMA AIRS AI Gateway sits between applications or agents and AI providers. It routes requests through a unified API and provides visibility into usage, cost and activity, with controls for model access, budgets, rate limits and guardrails.
Teams can use the SaaS service or host the gateway data plane in their own Kubernetes cluster. The official Portkey pricing page still lists Portkey plans, while Palo Alto Networks' current licensing documentation describes Prisma AIRS AI Gateway as consumption-based and licensed through Software NGFW credits; exact credit costs require a sales quote.
- Category
- Developer Tools
- Read from the maker’s pages
- 2026-11-19
Company
Features
- Route LLM requests through a unified API and switch providers without application code changes
- Configure fallback models, retries and load balancing
- Inspect prompts and responses inline for sensitive data and prompt injection
- Set workspace-level model access, budgets and rate limits
- Track requests, users, projects, token counts, latency and cost
- Deploy as a hosted service or host the data plane in your Kubernetes cluster
- Self-host the open-source gateway under the MIT License
Use cases
- Route an application’s model requests across providers
- Set model-access policies and spending limits for different teams
- Inspect AI requests and responses for security risks
- Track AI usage and costs across applications and projects
- Run the gateway data plane inside your own Kubernetes environment
Pricing
Free Forever
Free Forever
- 10k recorded logs per month
- 3 days for Logs, 30 days for metrics
- Universal API, Fallbacks, Loadbalancing, Retries & more
Production
$49/month
- 100k recorded logs per month
- +$9 overages per additional 100k requests
- LLM & Partner Guardrails
- Role-Based Access Control, Service Account API Keys
Enterprise
Custom Pricing
- Custom retention periods for Logs & Metrics
- SSO, granular budgets and rate limits
- Private Cloud Deployment and VPC Hosting
Capabilities
- Choice of models — “This gives you the freedom to swap AI vendors instantly without having to rewrite any of your internal applications.” source
- Self-hosted — “No. Prisma AIRS AI Gateway is deployed within your infrastructure, close to your agents and workloads, minimizing network overhead.” source
- API — “Portkey lets you access 1,600+ LLMs via a unified API, so you can focus on building, not managing.” source
- Official SDKs — “import Portkey from "portkey-ai" ;” source
- Traces and evaluates — “Monitor LLM behavior, catch anomalies early, and manage usage proactively with our real-time observability dashboard.” source
Latest updates
- New Integration: Hermes Agent + Portkey
Hermes Agent works with Portkey's OpenAI-compatible endpoint for model access, request logging, cost tracking, budget limits, and MCP governance.
- Agent Gateway - now live on Beta 🚀
Register agents and use a Portkey endpoint for governance, access control, observability, fallbacks, load balancing, security guardrails, and one auth model.
- Pi Agent is now on Portkey
Run Pi through Portkey for observability, budget and token limits, fallbacks, retries, caching, governance, and access controls.
- New: Portkey × Conductor integration
Use Portkey with Conductor for model access, fallbacks, load balancing, caching, routing, cost and token limits, logs, and metrics.
- Secret References
Use secret_mappings to fetch credentials from AWS Secrets Manager, Azure Key Vault, or HashiCorp Vault at runtime.
Where it works
Web · Linux · API
Alternatives to PRISMA AIRS AI Gateway
- Helicone
Helicone routes, logs, and analyzes AI application requests, helping teams debug model calls, track costs, and manage prompts.
- LiteLLM
LiteLLM routes requests to multiple LLM providers through one OpenAI-compatible API, with a Python SDK and a self-hosted gateway.
- Dataiku LLM Mesh
Dataiku LLM Mesh routes and governs access to hosted and self-hosted language models, with controls for usage, cost, performance, and safety.
- Langfuse
Langfuse traces AI applications, manages prompts, and evaluates model and agent outputs using cloud or self-hosted deployments.
- OpenRouter
OpenRouter provides one API to access and route requests across AI models from multiple providers, with model selection, fallbacks, and usage tracking.
- Portkey Guardrails
Portkey Guardrails checks AI inputs, outputs, and embedding requests, then can block, retry, or route requests based on results.
More from Portkey
- Portkey Guardrails
Portkey Guardrails checks AI inputs, outputs, and embedding requests, then can block, retry, or route requests based on results.
Questions about PRISMA AIRS AI Gateway
Is PRISMA AIRS AI Gateway free?
Yes. The maker’s pages state that PRISMA AIRS AI Gateway can be used for free. Read on the maker’s pages on 19 Nov 2026.
Does PRISMA AIRS AI Gateway have an API?
Yes. The maker’s pages state that PRISMA AIRS AI Gateway has an API. Read on the maker’s pages on 19 Nov 2026.
Which platforms does PRISMA AIRS AI Gateway run on?
PRISMA AIRS AI Gateway runs on Web, Linux and API, as its maker’s pages state. Read on the maker’s pages on 19 Nov 2026.
Is PRISMA AIRS AI Gateway open source?
Yes. The maker’s pages state that PRISMA AIRS AI Gateway is open source. Read on the maker’s pages on 19 Nov 2026.
Is PRISMA AIRS AI Gateway still available?
Yes, as of 19 Nov 2026: its maker’s pages were read that day and carried no notice that PRISMA AIRS AI Gateway has ended.
What changed recently in PRISMA AIRS AI Gateway?
The latest entry in its maker’s changelog: 29 Apr 2026, “New Integration: Hermes Agent + Portkey”. Before it: 21 Apr 2026, “Agent Gateway - now live on Beta 🚀”. Changelog checked on 19 Nov 2026.
Sources
- https://portkey.ai/pricing — Official pricing page; lists Free Forever, Production at $49/month, and Enterprise at Custom Pricing under Portkey.
- https://portkey.ai/features/ai-gateway — Official product page; says Portkey is now PRISMA AIRS AI Gateway and describes gateway capabilities.
- https://docs.paloaltonetworks.com/prisma-airs/ai-gateway/ai-gateway-overview — Official product documentation; describes routing, inline guardrails, workspaces and SaaS or hybrid deployment.
- https://docs.paloaltonetworks.com/prisma-airs/activation-and-onboarding/license-ai-gateway — Official licensing documentation; says the current gateway uses token metering and Software NGFW credits.
- https://www.paloaltonetworks.com/blog/2026/07/announcing-general-availability-of-prisma-airs-ai-gateway/ — Official announcement; gives the general-availability launch date and product capabilities.
- https://github.com/Portkey-AI/gateway/blob/main/LICENSE — Gateway repository license; identifies the MIT License.
Not confirmed
- Prices
- The published plan prices are on the Portkey-branded page; the current Prisma AIRS licensing documentation instead describes credit-based consumption, and does not give exact costs.
- Latest version
- No current product version or model name was identified on the product, pricing or release pages.
- Platforms
- Linux is inferred from the documented Kubernetes-hosted deployment; a specific supported operating-system list was not found.