Portkey
Also known as: Prisma AIRS AI Gateway
AI gateway, now presented as Prisma AIRS AI Gateway: one API to many model providers with routing, caching and guardrails, plus MCP and agent gateways, full request and tool-call logging, and enterprise governance.
Portkey is an AI gateway, a control plane between an organization's applications and agents and the models and tools they call. Palo Alto Networks acquired Portkey in 2026, and portkey.ai now announces that Portkey is Prisma AIRS AI Gateway, generally available for enterprises, while still selling Portkey's own plans.
The gateway puts many model providers behind one Universal API, so teams switch or mix models through routing configs instead of rewriting integrations, with fallbacks, retries, load balancing, conditional routing, circuit breakers and simple and semantic caching. The same configs run canary tests, sending a share of production traffic to a new model or prompt. Around the model traffic, the MCP Gateway gives agents one authenticated route to internal and external MCP servers, injecting credentials and checking per-user access for each tool, and the Agent Gateway proxies registered A2A agents with the same access control.
Every request and tool call is logged with its user, parameters, response, cost and latency, with traces, feedback scores and metadata, and guardrails check inputs and outputs and deny, flag, retry or fall back on a failed check.
Admins set budgets and rate limits per key, workspace or provider, and Enterprise adds SSO, SCIM, audit logs, logs and OpenTelemetry export, private cloud and VPC hosting, and SOC 2 Type 2 and HIPAA support. A free Developer plan and a $49 a month Production plan are self-serve, Enterprise is custom, and the open source gateway can be self-hosted free. Portkey governs and records agent traffic. It does not run agents, keep their memory or index the customer's documents.
Vendor details
Canonical URL
https://portkey.ai
Category
Agent infrastructure
Funding status
Part of Palo Alto Networks since its 2026 acquisition; portkey.ai presents Portkey as Prisma AIRS AI Gateway while still selling Portkey plans.
Company status
acquired
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
One Universal API across many model providers, including private deployments on AWS, GCP and Azure. The MCP Gateway fronts internal and external MCP servers with OAuth, bring-your-own auth, identity forwarding and JWT validation, and the Agent Gateway fronts A2A agents. Logs and analytics export through OpenTelemetry, SCIM works with Okta and Microsoft Entra, and partner guardrails plug into the gateway.
In practice
Your apps each integrate a different model provider, and switching is a rewrite every time. Portkey gives you one gateway in front of more than 1,600 models from many providers, so you route or swap models, with automatic fallbacks, without changing your code.
Your agents are quietly burning through tokens and you have no single view of the spend. Portkey logs every model and tool call with cost and latency, and enforces rate limits and budget guardrails to stop runaway usage.
Security needs control over agent traffic without slowing developers down. Portkey acts as a control plane that routes, monitors, and governs every AI request, flagging or blocking non-compliant behavior in real time.
Sources & related URLs
Agentic Index coverage score
7.5 / 14 capabilities · 54%
| Integrations & Tool Calling | Full |
|---|---|
|
Agents authenticate once to Portkey's MCP Gateway, which sits between them and MCP servers. For each tool call, the gateway checks that user's access to the server and tool, then injects the upstream credentials (OAuth, API keys or custom headers, with external identity providers such as Okta and Auth0). Servers and individual tools are provisioned per workspace and user. Through the gateway, agents take authenticated action in outside systems. Sourcedocs.portkey.ai/docs/product/mcp-gatewayread 2026-09-22 |
|
| Workflow Orchestration | Not documented |
|
The Agent Gateway is a proxy between agent clients and the customer's own agent servers. It handles authentication, access control and observability for registered A2A agents, and the agents and their steps run on the customer's servers. Portkey has no workflow definitions and runs no agents of its own, so nothing in it sequences or branches agent steps. Sourcedocs.portkey.ai/docs/product/agent-gatewayread 2026-09-22 |
|
| Knowledge Grounding & RAG | Not documented |
|
Embedding and file requests route through Portkey to providers, and responses are cached. Portkey maintains no retrieval structure over the customer's documents, and routing embedding calls builds none. Sourceportkey.ai/pricingread 2026-09-22 |
|
| Human Oversight & Guardrails | Partial |
|
Inputs and outputs pass through Portkey Guardrails, which act on the verdict automatically by denying the request, flagging it and letting it through, retrying or falling back. The MCP Gateway applies content filters per server, per user tool provisioning and rate limits on tool calls. Org and workspace admins can enforce guardrails, budgets and rate limits, so the customer sets these limits on what an agent may do. Portkey describes no step where a person reviews, approves or must validate an agent action before it commits. Sourcedocs.portkey.ai/docs/product/guardrailsread 2026-09-22 |
|
| Security, Identity & Governance | Full |
|
Enterprise plans get SSO through OIDC. Portkey also has fine grained role based access control, encrypted storage of provider keys, TLS 1.2+ in transit and AES-256 at rest. It claims compliance with SOC 2, ISO 27001, GDPR and HIPAA, with regular third party audits and penetration tests, and its Trust Center is at trust.portkey.ai. Enterprise includes SOC 2 Type 2 and custom BAAs. SCIM works with Okta and Microsoft Entra, and Portkey supports audit logs, JWT authentication and API key rotation. Sourcedocs.portkey.ai/docs/product/enterprise-offering/security-portkeyread 2026-09-22 |
|
| Observability & Auditability | Full |
|
Every tool call through the MCP Gateway is logged with the user, parameters and response, and the log can be filtered by server, user or time range. Model requests carry logs, traces, feedback, custom metadata, filters and alerts on every plan, with retention set per plan. Enterprise adds audit logs, logs export and OpenTelemetry export. Sourcedocs.portkey.ai/docs/product/mcp-gatewayread 2026-09-22 |
|
| Memory & State Persistence | Not documented |
|
Features cover routing, caching, observability, guardrails, prompt management and the MCP and agent gateways. None of them is a memory layer that keeps an agent's state or a user's context across sessions. The response cache holds model responses. Sourceportkey.ai/pricingread 2026-09-22 |
|
| Deployment & Data Residency | Full |
|
Enterprise offers private cloud deployment, VPC hosting and data isolation, and the open source gateway can be hosted by the customer. Hybrid deployments run on the customer's AWS (EKS, ECS, Marketplace), GCP and Azure (AKS, ACA), and air gapped deployment is available. A Gateway URLs setting points the dashboard at the customer's own self hosted AI and MCP gateways. Some of these options are Enterprise only. Sourceportkey.ai/pricingread 2026-09-22 |
|
| Prebuilt Agents, Templates & Packs | Not documented |
|
No ready made agents, workflows, templates or role specific agents ship with Portkey. The Prompt Library holds the customer's own templates, and the Skills Registry shares skills the team writes. The prebuilt assets are guardrail checks and partner guardrail integrations, which constrain what agents do. Sourcedocs.portkey.ai/docs/product/guardrailsread 2026-09-22 |
|
| Triggers & Channel Coverage | Not documented |
|
Work reaches agents through Portkey only when a client sends a request. Model, MCP and A2A calls are all made by the caller, and the Agent Gateway proxies them to the customer's agent servers. Portkey names no schedule, event, webhook or inbound queue that starts an agent run. The gateway waits for callers and does not wake an agent on its own. Sourcedocs.portkey.ai/docs/product/agent-gatewayread 2026-09-22 |
|
| Model Flexibility & Routing | Full |
|
Model choice sits with the customer. Portkey's Universal API puts many model providers behind one interface. Routing configs the customer writes set providers, model overrides, fallbacks, load balancing weights, retries, conditional routing and circuit breakers. Private model deployments on AWS, GCP and Azure are supported too. Sourceportkey.ai/pricingread 2026-09-22 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
Developers get Portkey's own API and SDKs. The same feedback call works through the Node and Python SDKs and the REST API, and Portkey publishes an OpenAPI specification. Prompts are served through a Prompt API, and the Agent Gateway acts as a server to downstream agent clients for registered A2A agents. Sourcedocs.portkey.ai/docs/product/observability/feedbackread 2026-09-22 |
|
| Testing, Debugging & Optimization | Full |
|
Canary testing sends a set share of production traffic to a new model or prompt through a routing config, on every plan. Each request carries guardrail verdicts (pass or fail per check, counted in the log drawer) and weighted feedback scores sent through the Feedback API, which can be analyzed by metadata. Guardrails can deny, retry or fall back on a failed check. Together these let a team compare a change against current traffic. Sourcedocs.portkey.ai/docs/product/ai-gateway/canary-testingread 2026-09-22 |
|
| Browser & Computer Use | Not documented |
|
The gateway routes model calls, including provider computer use tools, whose actions the customer's own code carries out. Portkey describes no browser, desktop or remote computer session that it runs for an agent, and it wires in no headless fetch or third party browser engine. Sourceportkey.ai/pricingread 2026-09-22 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
Upstream MCP servers that use the OAuth client credentials grant can now sit behind Portkey's MCP Gateway, starting with Enterprise Gateway 2.32.0. Admins enter the client ID, client secret, token endpoint and an optional scope, and the gateway requests only that scope and fetches and refreshes tokens on its own. The same release lets the Agent Gateway pass chosen headers from the caller's request through to the upstream agent.
Bears on: MCP / tool calling / API
View sourcePricing
Free Developer plan · Production $49/mo · open source gateway free to self-host
usage
Included quota
Dev: 10K recorded logs a month with basic observability and gateway features. Production: 100K logs with $9 per additional 100K (up to 3M), 30 day retention, guardrails, RBAC, semantic caching. Enterprise: 10M+ logs, custom retention, SSO, private cloud/hybrid/airgapped deployment.
What is public
Portkey publishes self-serve pricing: a free Developer plan, Production at $49 a month with a published overage, and custom Enterprise with private cloud, VPC and data isolation options. The open source gateway self-hosts free with no request limit.
Billing mechanics
Flat plan fee bundles a monthly recorded log allowance; additional logs bill in $9 per 100K increments. Model/provider costs are the customer's own since Portkey is a routing layer.
Cost watchouts
Billing is on recorded logs: requests keep flowing past the plan's log allowance, but logs beyond it are not recorded, which leaves traffic unobserved on an undersized plan. Log retention is 3 days on Developer and 30 days on Production; longer retention, SSO and data residency options are Enterprise only.
Variable cost rationale
Log volume metering scales with traffic, but the $9 per 100K overage is modest; the bigger swing is the jump to Enterprise for retention and compliance.
Additional watchouts
Buyers outside the Palo Alto Networks ecosystem should weigh how Prisma AIRS integration priorities shape the standalone product roadmap.
Overage / add-ons
Production: $9 for every additional 100K requests, up to 3M a month; Developer: no overage, logs past 10K are not recorded
Sales call required
Mixed (some tiers require a call)
Free / trial
Developer plan free forever (10K recorded logs a month); open source gateway free to self-host
Lowest paid plan
Production $49/mo (100K recorded logs, $9 per additional 100K)
Commercial notes
Acquired by Palo Alto Networks (announced Apr 30, 2026, closed May 29, 2026) to serve as the AI Gateway for Prisma AIRS. Prior: $15M Series A (Feb 2026, Elevation/Lightspeed). Routes to 1,600+ models across providers; 2025 Gartner Cool Vendor in LLM Observability.
Key ambiguities
Enterprise pricing is unpublished. portkey.ai now announces that Portkey is Prisma AIRS AI Gateway under Palo Alto Networks while still selling the Portkey plans above; packaging may change.
Missing data
Enterprise pricing unpublished; post acquisition plan roadmap unconfirmed.
Related vendors
- AgentOps — Agent observability and debugging platform: open source SDKs trace…
- AgentPhone — AgentPhone gives the AI agents a team builds their own US or…
- Agno — Python agent framework and AgentOS runtime (formerly Phidata) for…
- AIsa — Resource and payment gateway for AI agents, with one key to 110+…
- Albato — Albato is a no code integration platform that sells an embedded,…
- AlphaBitCore — AI control plane for regulated financial firms: one gateway enforces…
Alternatives to Portkey
The closest documented capability profiles to Portkey among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Arcade7.5 / 14Matches Portkey across all 14 documented capabilities
- Kong8.5 / 14Adds documented Knowledge Grounding & RAG
- Freeplay8.0 / 14Adds documented Triggers & Channel Coverage
- Metorial8.0 / 14Adds documented Triggers & Channel Coverage
- Stacklok8.0 / 14Adds documented Prebuilt Agents, Templates & Packs
- Clawvisor6.5 / 14Fuller documented coverage on Human Oversight & Guardrails
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded