Portkey AI Gateway: Pricing, Pros & Cons

Founded 2023|San Francisco, California|Freemium

What is Portkey?

Portkey is an AI gateway with observability, guardrails, and prompt management, used to route LLM requests across providers through one OpenAI-compatible API. The gateway handles retries, fallbacks, load balancing, and caching, and its core is open source and self-hostable.

Palo Alto Networks completed its acquisition of Portkey on May 29, 2026, and the product is now the Prisma AIRS AI Gateway.

The hosted plans meter recorded logs rather than requests. The free Developer plan records 10,000 logs a month with three-day retention, and Production costs $49 a month for 100,000 logs plus $9 per additional 100,000, with 30-day log retention. Requests past the free limit still go through, but they aren't logged, so that traffic has no record to debug from. If the acquisition changes your roadmap assumptions, the Portkey alternatives worth comparing include platforms that pair a gateway with tracing and automated evals.

Key features

Core capabilities this platform advertises.

Unified API across major providers
Retries, fallbacks, and load balancing
Simple and semantic caching
Deterministic and LLM guardrails
Request logs with cost and latency analytics

Strengths and tradeoffs

What this tool does well, and the limitations to keep in mind.

Pros

Open-source, self-hostable gateway
Routing, caching, and guardrails in one gateway
Requests keep flowing past the free log limit
RBAC and semantic caching on the $49 Production plan

Cons

Now part of Palo Alto Networks as Prisma AIRS AI Gateway
Three-day log retention on the free plan
Requests past the log cap go unrecorded on the free plan
Semantic caching and LLM guardrails require Production

Plans & pricing

What's included in each plan, and how the tiers compare.

Developer

Free

10K recorded logs/mo
3-day log retention
3 prompt templates
Simple caching and deterministic guardrails

Production

$49

Per month

100K recorded logs/mo, then $9 per 100K
30-day log retention
Unlimited prompt templates
Semantic caching and LLM guardrails

Enterprise

Custom

Per month

10M+ recorded logs/mo
VPC hosting or private cloud
SSO with Okta
Custom retention and data export

Open source

Free

Self-hosted

Universal API and routing
Load balancing and guardrails

View official pricing page

Common use cases

Engineering teams that need a gateway with routing, caching, and guardrails for production LLM apps

Multi-provider failover and load balancing
Cutting cost through caching
Guardrails on production traffic
Centralized LLM access management

Using Portkey with Respan

Respan traces Portkey SDK apps through respan-instrumentation-portkey, which normalizes Portkey's spans into Respan's tracing pipeline. To move routing as well, point any OpenAI-compatible client at the Respan gateway for the same flow, with request logs, fallbacks, prompt management, and automated evals on every call across 1,000+ models.

Trace existing Portkey SDK calls with one instrumentor
Switch to Respan's OpenAI-compatible gateway by changing the base URL
Score output on live production traffic with automated evals
Route across 1,000+ models with a 99.9% uptime SLA on Team

Switch gateways by changing one base URL

10KFree traces/mo

1,000+Models

5 minSetup

Respan's OpenAI-compatible gateway routes across 1,000+ models with fallbacks, traces every request, and scores output on live traffic. Start free.

Try Respan free

Best Portkey alternatives & competitors

Top companies in LLM Gateways you can use instead of Portkey.

Respan

One API for 1,000+ models