Compare tools
Pick 2–5 published tools. Values come from our evidence-checked profiles — an em-dash means unknown, never “no”. Ranking rows reflect the current published methodology for the comparison context.
Portkey vs Kong AI Gateway
Context: LLM Gateways & Model Routing Platforms · 21 rows
| Attribute | PortkeyPortkey | Kong AI GatewayKong |
|---|---|---|
| Identity · Taxonomy | ||
| Documentation | https://portkey.ai/docs | https://developer.konghq.com/ai-gateway/ |
| Tags | llm-gateway, llm-observability, guardrails, prompt-management | llm-gateway, api-gateway, kubernetes, mcp, microservices |
| Platform · Deployment | ||
| Delivery model | SaaS, self-hosted | SaaS, self-hosted |
| Pricing · Licensing | ||
| Free plan | true | — |
| License model | open-source | open-source |
| Pricing model | free, subscription, custom | custom |
| Source repository | https://github.com/Portkey-AI/gateway | https://github.com/Kong/kong |
| SPDX license IDs | MIT | Apache-2.0 |
| Starting price | 49 | — |
| Features · Product · Fit | ||
| Core capabilities | Unified API to 1,600+ LLMs, Routing, fallbacks, and load balancing, Observability with logs, traces, and metrics, Guardrails including PII redaction, Prompt management and caching, MCP gateway for tool access | Multi-LLM routing with provider failover, Semantic caching and load balancing, PII sanitization and semantic prompt guards, Token quotas, chargeback, and spend tracking, MCP server governance and authentication, L7 observability with logging and tracing |
| Api · Integrations · Ecosystem | ||
| API available | true | true |
| Security · Privacy · Compliance | ||
| Regulatory compliance | GDPR, HIPAA | — |
| Security certifications | SOC 2 Type II, HIPAA | — |
| Support · Docs · Services | ||
| Support channels | community, production support, priority support | — |
| Audience · Use · Case | ||
| Best for | Teams building GenAI applications that want observability, guardrails, and gateway routing available as SaaS or self-hosted. | Organizations with existing API gateway or Kubernetes infrastructure that want LLM, MCP, and agent traffic governed through one gateway platform. |
| Company-size fit | SMB, mid-market, enterprise | SMB, mid-market, enterprise |
| Primary use cases | Production monitoring for GenAI applications, LLM cost and latency management, AI governance with RBAC and audit logs | Governing LLM traffic in API infrastructure, MCP and agent-to-agent traffic governance, Provider switching and failover |
| Skill level | intermediate | advanced |
| Lifecycle · Versioning | ||
| Maintenance status | active | active |
| Vendor · Maintainer | ||
| Headquarters country | United States | — |
| Vendor type | commercial | commercial |
“—” means unknown (we never treat missing evidence as a negative). Cells reflect each value's current evidence state; superseded or stale values are excluded. Ranking rows are editorial assessments under the cited methodology version — see the LLM Gateways & Model Routing Platforms category for definitions.