Compare AI gateways

Pick up to four platforms. Every cell shows a verification state, not a marketing claim — “not publicly documented” means we could not find an official source, not that the capability is missing. See our methodology.

AI gateway feature comparison
CapabilityPKPortkey
Security
Prompt-injection detectionIdentifies attempts to override system instructions in prompts or tool output.Verification pendingVerification pending. Verification pending.
Prompt-injection blockingBlocks the request before it reaches the model, rather than only logging it.Verification pendingVerification pending. Verification pending.
Jailbreak detectionDetects known jailbreak and guardrail-evasion patterns.Verification pendingVerification pending. Verification pending.
PII detectionIdentifies personal information in prompts or responses.AvailableAvailable. Verified from official documentation.Available as a guardrail check.
PII redactionMasks or removes personal information before it leaves the gateway.LimitedLimited. Partially documented.Redaction is available through guardrail actions; coverage varies by check.
Secret & credential scanningDetects API keys, tokens and credentials in prompts or output.Verification pendingVerification pending. Verification pending.
Input inspectionInspects outbound requests before they reach the model.AvailableAvailable. Verified from official documentation.Input guardrails run before the request reaches a model.
Output inspectionInspects inbound model responses before they reach the application.AvailableAvailable. Verified from official documentation.Output guardrails run on model responses.
Content moderationClassifies harmful or policy-violating content. Not equivalent to injection defense.AvailableAvailable. Verified from official documentation.
Policy enforcementConfigurable rules applied per route, team, key or environment.AvailableAvailable. Verified from official documentation.Guardrails are configured per route, key or workspace.
Agent & tool controlsConstrains tool calls, autonomous actions and agent permissions.Verification pendingVerification pending. Verification pending.
MCP securityControls or inspects Model Context Protocol server traffic.AvailableAvailable. Verified from official documentation.MCP gateway support.
Audit logsDurable, reviewable record of requests and policy decisions.AvailableAvailable. Verified from official documentation.Request logs and traces.
Access controlsRoles, teams, keys and scoped permissions.AvailableAvailable. Verified from official documentation.Workspaces, virtual keys and role-based access.
Cost optimization
Prompt compressionReduces the number of tokens actually sent to the model.Not publicly documentedNot publicly documented. Not publicly documented.
Exact cachingReturns a stored response for a byte-identical request.AvailableAvailable. Verified from official documentation.
Semantic cachingReturns a stored response for a semantically similar request using embeddings.AvailableAvailable. Verified from official documentation.
Intelligent model routingChooses a model per request based on task, cost or performance.AvailableAvailable. Verified from official documentation.Conditional routing across providers and models.
Automatic fallbackRetries against an alternate model or provider on failure.AvailableAvailable. Verified from official documentation.
Load balancingDistributes traffic across keys, regions or providers.AvailableAvailable. Verified from official documentation.
Token budgetsHard or soft limits expressed in tokens.AvailableAvailable. Verified from official documentation.Budget limits can be expressed in tokens per key.
Dollar budgetsSpending limits expressed in currency.AvailableAvailable. Verified from official documentation.Cost budgets per virtual key or workspace.
Rate limitsRequest throttling. Throttling is not the same as token reduction.AvailableAvailable. Verified from official documentation.
Cost attributionAttributes spend to teams, keys, users or features.AvailableAvailable. Verified from official documentation.Spend is attributed per key, workspace and metadata tag.
Cost analyticsDashboards and reporting on token and dollar usage.AvailableAvailable. Verified from official documentation.
Savings reportingReports the spend avoided by compression, caching or routing.Verification pendingVerification pending. Verification pending.
Models and deployment
OpenAISupports OpenAI models.AvailableAvailable. Verified from official documentation.
AnthropicSupports Anthropic models.AvailableAvailable. Verified from official documentation.
Google GeminiSupports Google Gemini models.AvailableAvailable. Verified from official documentation.
AWS BedrockSupports AWS Bedrock as a provider.AvailableAvailable. Verified from official documentation.
Azure OpenAISupports Azure OpenAI deployments.AvailableAvailable. Verified from official documentation.
Open-source modelsSupports open-weight models or self-hosted inference endpoints.AvailableAvailable. Verified from official documentation.
SaaS deploymentVendor-hosted managed service.AvailableAvailable. Verified from official documentation.
VPC / private cloudDeployable inside a customer cloud account.AvailableAvailable. Verified from official documentation.Private-cloud deployment for enterprise plans.
On-premiseDeployable in a customer data centre.Verification pendingVerification pending. Verification pending.
Self-hostedCustomer runs and operates the gateway themselves.AvailableAvailable. Verified from official documentation.The gateway can be run from the open-source repository.
Open-source optionSource code published under an open-source licence.AvailableAvailable. Verified from official documentation.The gateway is published as a public repository.
APIProgrammatic API for configuration or traffic.AvailableAvailable. Verified from official documentation.
SDKOfficial client libraries.AvailableAvailable. Verified from official documentation.Official Python and Node SDKs.
Commercial
Public pricingList pricing published on the vendor website.Verification pendingVerification pending. Verification pending.
Free tierA permanently free usage tier.Verification pendingVerification pending. Verification pending.
Free trialA time-limited or credit-limited trial.Verification pendingVerification pending. Verification pending.
Enterprise planA named enterprise tier.Verification pendingVerification pending. Verification pending.
Contact sales requiredPricing requires contacting the vendor.Verification pendingVerification pending. Verification pending.
Best forTeams that want an open-source gateway with guardrails and spend controls
Last reviewed2026-07-30