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
CapabilityCGConstellation Gate AI
Security
Prompt-injection detectionIdentifies attempts to override system instructions in prompts or tool output.AvailableAvailable. Verified from official product page.Product page states attacks are screened before reaching the model.
Prompt-injection blockingBlocks the request before it reaches the model, rather than only logging it.AvailableAvailable. Verified from official product page.Product page shows a "Blocked · Injection" state on outbound requests.
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 product page.
PII redactionMasks or removes personal information before it leaves the gateway.AvailableAvailable. Verified from official product page.Product page shows a "Redacted · PII" state.
Secret & credential scanningDetects API keys, tokens and credentials in prompts or output.AvailableAvailable. Verified from official product page.Product page shows a "Flagged · Secret" state for credential patterns.
Input inspectionInspects outbound requests before they reach the model.AvailableAvailable. Verified from official product page.Outbound requests are described as inspected inline.
Output inspectionInspects inbound model responses before they reach the application.AvailableAvailable. Verified from official product page.Inbound responses are described as inspected inline.
Content moderationClassifies harmful or policy-violating content. Not equivalent to injection defense.Verification pendingVerification pending. Verification pending.
Policy enforcementConfigurable rules applied per route, team, key or environment.Verification pendingVerification pending. Verification pending.
Agent & tool controlsConstrains tool calls, autonomous actions and agent permissions.LimitedLimited. Partially documented.Hijacked tool calls are described as stopped; the configuration surface is not yet verified.
MCP securityControls or inspects Model Context Protocol server traffic.Verification pendingVerification pending. Verification pending.
Audit logsDurable, reviewable record of requests and policy decisions.Verification pendingVerification pending. Verification pending.
Access controlsRoles, teams, keys and scoped permissions.Verification pendingVerification pending. Verification pending.
Cost optimization
Prompt compressionReduces the number of tokens actually sent to the model.AvailableAvailable. Verified from official product page.Product page shows a "Compressed" state with a percentage prompt reduction.
Exact cachingReturns a stored response for a byte-identical request.Verification pendingVerification pending. Verification pending.
Semantic cachingReturns a stored response for a semantically similar request using embeddings.Verification pendingVerification pending. Verification pending.
Intelligent model routingChooses a model per request based on task, cost or performance.LimitedLimited. Partially documented.One endpoint fronting many models is advertised; cost-based routing is not documented.
Automatic fallbackRetries against an alternate model or provider on failure.Verification pendingVerification pending. Verification pending.
Load balancingDistributes traffic across keys, regions or providers.Verification pendingVerification pending. Verification pending.
Token budgetsHard or soft limits expressed in tokens.Verification pendingVerification pending. Verification pending.
Dollar budgetsSpending limits expressed in currency.Verification pendingVerification pending. Verification pending.
Rate limitsRequest throttling. Throttling is not the same as token reduction.Verification pendingVerification pending. Verification pending.
Cost attributionAttributes spend to teams, keys, users or features.Verification pendingVerification pending. Verification pending.
Cost analyticsDashboards and reporting on token and dollar usage.Verification pendingVerification pending. Verification pending.
Savings reportingReports the spend avoided by compression, caching or routing.LimitedLimited. Partially documented.Token spend reduction is advertised; a savings report surface is not verified.
Models and deployment
OpenAISupports OpenAI models.AvailableAvailable. Verified from official product page.
AnthropicSupports Anthropic models.AvailableAvailable. Verified from official product page.
Google GeminiSupports Google Gemini models.AvailableAvailable. Verified from official product page.
AWS BedrockSupports AWS Bedrock as a provider.Verification pendingVerification pending. Verification pending.
Azure OpenAISupports Azure OpenAI deployments.Verification pendingVerification pending. Verification pending.
Open-source modelsSupports open-weight models or self-hosted inference endpoints.LimitedLimited. Partially documented.OpenRouter is listed as a reachable model API.
SaaS deploymentVendor-hosted managed service.AvailableAvailable. Verified from official product page.
VPC / private cloudDeployable inside a customer cloud account.Verification pendingVerification pending. Verification pending.
On-premiseDeployable in a customer data centre.Verification pendingVerification pending. Verification pending.
Self-hostedCustomer runs and operates the gateway themselves.Verification pendingVerification pending. Verification pending.
Open-source optionSource code published under an open-source licence.Verification pendingVerification pending. Verification pending.
APIProgrammatic API for configuration or traffic.AvailableAvailable. Verified from official product page.One OpenAI-style endpoint fronting multiple model APIs.
SDKOfficial client libraries.Verification pendingVerification pending. Verification pending.
Commercial
Public pricingList pricing published on the vendor website.Verification pendingVerification pending. Verification pending.
Free tierA permanently free usage tier.AvailableAvailable. Verified from official product page.Free tier included, no credit card required.
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 running coding agents that want prompt security and token reduction in one hop
Last reviewed2026-07-30