# AI gateway provider directory Published by OneQuill, developer of OneVir. This is a documentary review of selected public sources, not an independent vendor security audit. Product coverage is a researched snapshot, not an exhaustive inventory. Category and deployment mode describe the cited offering; editions, contracts and configuration can change the scope. Selected cases illustrate evaluation questions. Advisory counts are not security rankings. Sources reviewed: 2026-10-07 ## agentgateway Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Rust gateway for AI and MCP. Local rate limits are held in memory; a remote rate-limit service supplies a different sharing boundary. Evaluation question: Who can author policies and reference credentials across namespaces? - [Product / project](https://agentgateway.dev/) - [Documentation / policy](https://agentgateway.dev/docs/standalone/latest/documentation/configuration/resiliency/rate-limits/) - [Related case](https://onequill.dev/resources/agentgateway-namespace-isolation) ## AI/ML API Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hosted model API. Service key management does not replace review of provider processing and retention terms. Evaluation question: Can keys be scoped and revoked without interrupting all applications? - [Product / project](https://aimlapi.com/) - [Documentation / policy](https://docs.aimlapi.com/api-references/service-endpoints/api-key-management) ## AIHubMix Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hosted model aggregation. Review the service policy alongside the actual selected upstream. Evaluation question: What provider, geography and content-retention terms govern this model? - [Product / project](https://aihubmix.com/developers) - [Documentation / policy](https://docs.aihubmix.com/en/terms-and-privacy/Privacy) ## Apache APISIX Category: API platforms. Deployment: Self-hosted. Reviewed: 2026-10-07. AI plugins extend the API gateway. Documented streaming timeouts can close a stream without a DONE event. Evaluation question: How does the client detect an incomplete stream and avoid unsafe replay? - [Product / project](https://apisix.apache.org/ai-gateway/) - [Documentation / policy](https://apisix.apache.org/docs/apisix/plugins/ai-proxy/) ## Apigee / Model Armor Category: Cloud gateways. Deployment: Managed. Reviewed: 2026-10-07. Model Armor is a configured integration. Inspection limits, supported regions and service latency affect coverage. Evaluation question: What is the explicit fail-open or fail-closed policy for inspection errors? - [Product / project](https://docs.cloud.google.com/model-armor/model-armor-apigee-integration) - [Documentation / policy](https://docs.cloud.google.com/apigee/docs/api-platform/tutorials/using-model-armor-policies) ## AWS Bedrock AgentCore Gateway Category: Cloud gateways. Deployment: Managed. Reviewed: 2026-10-07. AgentCore Gateway now documents inference targets as well as tool integration. Policy-language and target support still have limits. Evaluation question: Which inference and tool targets are in scope for each policy? - [Product / project](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway.html) - [Documentation / policy](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway-targets-inference.html) ## Azure API Management Category: Cloud gateways. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. Token-limit counters are independent across specified gateway, region and workspace boundaries. A newer AI gateway preview has separate limitations. Evaluation question: Is your budget global, or is it enforced independently per region? - [Product / project](https://learn.microsoft.com/en-us/azure/api-management/genai-gateway-capabilities) - [Documentation / policy](https://learn.microsoft.com/en-us/azure/api-management/llm-token-limit-policy) ## Bifrost Category: Specialist gateways. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. Gateway with MCP and custom-plugin administration. The September advisories distinguish dynamically linked builds from published static Docker images. Evaluation question: Is the management API authenticated and isolated from application callers? - [Product / project](https://www.getbifrost.ai/) - [Documentation / policy](https://docs.getbifrost.ai/) - [Related case](https://onequill.dev/resources/bifrost-management-api-boundaries) ## Braintrust AI Proxy Category: Specialist gateways. Deployment: Managed, Customer-hosted. Reviewed: 2026-10-07. Proxy and observability sit within Braintrust's architecture. Data-plane hosting and logging choices need to be assessed together. Evaluation question: Where are request logs and evaluation datasets stored? - [Product / project](https://www.braintrust.dev/) - [Documentation / policy](https://www.braintrust.dev/docs/platform/architecture) ## Cloudflare AI Gateway Category: Cloud gateways. Deployment: Managed. Reviewed: 2026-10-07. Account-level AI Gateway permissions can cover all gateways, including stored BYOK credentials. Logging behaviour changed for new customers from 24 September 2026. Evaluation question: How are tenants separated and which logging generation applies to your account? - [Product / project](https://developers.cloudflare.com/ai-gateway/) - [Documentation / policy](https://developers.cloudflare.com/ai-gateway/configuration/authentication/) ## CometAPI Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hosted model aggregator. Published privacy claims need a deployment-specific contractual scope. Evaluation question: Which upstreams receive content and which contractual exceptions apply? - [Product / project](https://www.cometapi.com/about/) - [Documentation / policy](https://www.cometapi.com/privacy-policy/) ## Databricks AI Gateway Category: Cloud gateways. Deployment: Managed. Reviewed: 2026-10-07. Unity-oriented gateway services and legacy Mosaic endpoints have different controls. Inference-table payload capture is a separate configuration. Evaluation question: Which service generation and payload-capture settings are enabled? - [Product / project](https://docs.databricks.com/aws/en/ai-gateway/) - [Documentation / policy](https://docs.databricks.com/gcp/en/ai-gateway/model-services) ## Eden AI Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hosted aggregation across AI services. The data-processing agreement and provider exceptions matter alongside security descriptions. Evaluation question: Does the DPA cover each selected service and subprocessor? - [Product / project](https://www.edenai.co/) - [Documentation / policy](https://www.edenai.co/dpa) ## Envoy AI Gateway / Agent Router Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Project naming and repository have moved towards Agent Router. Advisories retain the Envoy AI Gateway name and component scope. Evaluation question: Which release and MCP request-processing path are actually deployed? - [Product / project](https://aigateway.envoyproxy.io/) - [Documentation / policy](https://github.com/theagentrouter/agent-router) - [Related case](https://onequill.dev/resources/envoy-ai-gateway-mcp-message-smuggling) - [Related case](https://onequill.dev/resources/envoy-ai-gateway-mcp-request-limits) ## F5 NGINX Gateway Fabric Category: API platforms. Deployment: Self-hosted. Reviewed: 2026-10-07. AI Guardrails integration is a configured stack feature, not a property of every NGINX proxy. Verify policy acceptance and request-size handling. Evaluation question: Which guardrail integration and enforcement status are active? - [Product / project](https://docs.nginx.com/nginx-gateway-fabric/) - [Documentation / policy](https://docs.nginx.com/nginx-gateway-fabric/how-to/f5-ai-guardrails/) ## FastRouter Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hosted router with routing and privacy descriptions. Confirm the effective configuration rather than infer it from feature names. Evaluation question: What happens to data and cost when a route falls back? - [Product / project](https://fastrouter.ai/features) - [Documentation / policy](https://fastrouter.ai/privacy) ## GitLab AI Gateway Category: Related products. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. Application-specific gateway for GitLab Duo, not a general interchangeable model router. Hosted and self-hosted patch responsibilities differ. Evaluation question: Which Duo access and flow-template capabilities are exposed? - [Product / project](https://docs.gitlab.com/administration/gitlab_duo/) - [Documentation / policy](https://docs.gitlab.com/releases/patches/other-patches/patch-release-gitlab-ai-gateway-19-4-1-released/) - [Related case](https://onequill.dev/resources/gitlab-ai-gateway-template-sandbox) ## Gravitee Category: API platforms. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. AI policies and masking are edition and flow-order dependent. Payload logging introduces its own memory and retention needs. Evaluation question: Is masking applied before every logger and export in this edition? - [Product / project](https://www.gravitee.io/platform/ai-gateway) - [Documentation / policy](https://documentation.gravitee.io/apim/create-and-configure-apis/apply-policies/policy-reference/data-logging-masking) ## Helicone Category: Specialist gateways. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. Observability can capture requests automatically. Retention and self-hosting depend on the selected service and configuration. Evaluation question: Which content fields are logged, for how long, and under whose account? - [Product / project](https://www.helicone.ai/) - [Documentation / policy](https://docs.helicone.ai/getting-started/quick-start) ## Higress Category: API platforms. Deployment: Self-hosted. Reviewed: 2026-10-07. Gateway with AI provider and security-guard plugins. Plugin selection and policy order define the effective behaviour. Evaluation question: Which request and response paths are covered by the selected plugins? - [Product / project](https://higress.io/en/ai-gateway) - [Documentation / policy](https://github.com/higress-group/higress/security) ## Hugging Face Inference Providers Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hugging Face's own request handling and metadata retention differ from upstream provider policies. Dedicated Inference Endpoints are a separate service. Evaluation question: Which provider receives the request, and are its retention terms approved? - [Product / project](https://huggingface.co/docs/inference-providers/en/index) - [Documentation / policy](https://huggingface.co/docs/inference-providers/en/security) ## IBM API Connect Category: API platforms. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. AI support and limits vary between SaaS and software versions, including watsonx integrations. Evaluation question: Which AI features and limitations apply to your deployment and release? - [Product / project](https://www.ibm.com/docs/en/api-connect/cloud/saas?topic=applications-using-ai-gateway-support-watsonxai-apis) - [Documentation / policy](https://www.ibm.com/docs/en/api-connect/software/12.1.1?topic=overview-known-limitations) ## kgateway Category: API platforms. Deployment: Self-hosted. Reviewed: 2026-10-07. Kubernetes gateway with AI extensions. Envoy-based documentation is versioned; prompt guards and rate limits need configured policies. Evaluation question: What happens if an external guard or rate-limit service is unavailable? - [Product / project](https://kgateway.dev/) - [Documentation / policy](https://kgateway.dev/docs/envoy/2.1.x/ai/about/) ## Kong AI Gateway Category: API platforms. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. Traditional AI Proxy plugins and newer AI Gateway offerings have distinct version and deployment scopes. Evaluation question: How is streaming usage reconciled against provider invoices? - [Product / project](https://developer.konghq.com/index/ai-gateway/) - [Documentation / policy](https://developer.konghq.com/plugins/ai-proxy/changelog/) - [Related case](https://onequill.dev/resources/kong-gemini-streaming-token-accounting) ## LangDB Category: Specialist gateways. Deployment: Managed, Customer-hosted. Reviewed: 2026-10-07. AI gateway with observability. Documented ClickHouse TTL deletion runs through asynchronous merges. Evaluation question: What deletion latency applies to logs, backups and exports? - [Product / project](https://langdb.ai/) - [Documentation / policy](https://docs.langdb.ai/enterprise/resources/configuring-data-retention/) ## Leanroute Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Policy dated 1 October 2026 distinguishes database storage from response and semantic caches, which default to 15 minutes; no-persistence is a separate choice. Evaluation question: Are response caching and semantic caching disabled for sensitive routes? - [Product / project](https://leanroute.dev/) - [Documentation / policy](https://leanroute.dev/privacy) ## LiteLLM Category: Specialist gateways. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. Python proxy and SDK distribution are distinct from the vendor's pinned proxy Docker distribution. Evaluation question: What package or image digest is running, and how are credentials rotated after an incident? - [Product / project](https://www.litellm.ai/) - [Documentation / policy](https://docs.litellm.ai/docs/proxy/prod) - [Related case](https://onequill.dev/resources/litellm-march-2026-package-incident) ## LLM Gateway Category: Hosted routers. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. Metadata-only logging is described as the default; payload logging is an opt-in setting. Upstream retention remains separate. Evaluation question: Which content logging options and upstream terms are enabled? - [Product / project](https://llmgateway.io/) - [Documentation / policy](https://llmgateway.io/legal/privacy) ## LM Studio Category: Inference engines. Deployment: Self-hosted. Reviewed: 2026-10-07. Desktop and headless model runtime with an API server. Authentication is optional; network serving and LM Link alter the access and execution boundary. Evaluation question: Are API-token permissions enforced, which machine executes each model, and what model-loading and eviction behaviour does the application rely on? - [Product](https://lmstudio.ai/) - [API-token authentication](https://lmstudio.ai/docs/developer/core/authentication) - [Network server settings](https://lmstudio.ai/docs/developer/core/server/serve-on-network) - [LM Link remote execution](https://lmstudio.ai/docs/developer/core/lmlink) - [Offline operation](https://lmstudio.ai/docs/app/offline) - [Related case](https://onequill.dev/resources/lm-studio-network-authentication-and-lifecycle) ## Lunar.dev Category: Specialist gateways. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. Gateway and traffic-management offering. Production logs and data-plane placement need their own review. Evaluation question: Which data leaves the gateway through logs or management integrations? - [Product / project](https://www.lunar.dev/product/ai-gateway) - [Documentation / policy](https://docs.lunar.dev/api-gateway/lunar-dev-in-production/lunar-logs) ## Martian Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Hosted routing gateway. Authentication documentation establishes an API boundary, not every enterprise data-control claim. Evaluation question: What route decision evidence and data-processing terms can be supplied? - [Product / project](https://withmartian.com/) - [Documentation / policy](https://gateway-docs.withmartian.com/api-reference/authentication) ## MLflow AI Gateway Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Current gateway documentation distinguishes development credential-encryption defaults from production key management and SQL-backed rotation. Evaluation question: Is a production encryption secret configured and is rotation rehearsed? - [Product / project](https://mlflow.org/docs/latest/genai/governance/ai-gateway) - [Documentation / policy](https://www.mlflow.org/docs/latest/genai/governance/ai-gateway/api-keys/key-rotation/) ## New API Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Project with quota and billing functions. Release-candidate version scope matters in the quota-overflow advisory. Evaluation question: Can credits, reservations and final charges be reconciled under concurrent requests? - [Product / project](https://docs.newapi.pro/) - [Documentation / policy](https://docs.newapi.pro/en/docs/guide/wiki/changelog) - [Related case](https://onequill.dev/resources/new-api-quota-billing-overflow) ## nexos.ai Category: Specialist gateways. Deployment: Managed. Reviewed: 2026-10-07. Hosted gateway and AI platform. Procurement needs the service contract, provider list and configured routing policy. Evaluation question: What geography and retention commitments cover every fallback? - [Product / project](https://nexos.ai/ai-gateway/) - [Documentation / policy](https://nexos.ai/legal/privacy-policy/) ## Not Diamond Category: Related products. Deployment: Managed, Customer-hosted. Reviewed: 2026-10-07. Routing and model-selection API. It is listed as a related product rather than assumed to implement a full gateway boundary. Evaluation question: Which controls reside in your calling application and which in the routing service? - [Product / project](https://docs.notdiamond.ai/docs/what-is-not-diamond) - [Documentation / policy](https://docs.notdiamond.ai/docs/privacy-security-and-local-deployments) ## Ollama Category: Inference engines. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. Local model runtime and API, with optional cloud-model features. The local API does not require authentication; cloud API credentials and cloud processing are a separate mode. Evaluation question: Is execution local or cloud, who can reach inference and model-management endpoints, and how do context, parallel requests and model residency fit memory? - [Project](https://ollama.com/) - [Local and cloud API authentication](https://docs.ollama.com/api/authentication) - [Network, cloud and concurrency settings](https://docs.ollama.com/faq) - [Context and memory](https://docs.ollama.com/context-length) - [Related case](https://onequill.dev/resources/ollama-model-import-memory-boundary) ## One API Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Separate project from New API. Model mappings and protocol reconstruction can affect unsupported fields. Evaluation question: Which request fields survive translation, including tools and usage options? - [Product / project](https://github.com/songquanpeng/one-api) - [Documentation / policy](https://github.com/songquanpeng/one-api) ## OneVir Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. OneQuill's own product combines gateway controls and local execution paths. This profile is affiliated and is not an independent security assessment. Evaluation question: Which controls are enabled in your installed version, and which remain the inference backend's responsibility? - [Product / project](https://onevir.onequill.dev/) - [Documentation / policy](https://onevir.onequill.dev/#capabilities) ## OpenRouter Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Provider routing, fallback and endpoint-level privacy controls need to be configured together. ZDR and geography are route properties. Evaluation question: Can every selected endpoint and fallback satisfy the same privacy constraints? - [Product / project](https://openrouter.ai/) - [Documentation / policy](https://openrouter.ai/docs/guides/features/guardrails/overview) ## OpenZiti LLM Gateway Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Identity-based overlay networking is relevant to gateway access. Platform network features do not establish every application-level safeguard. Evaluation question: How are service identities issued, revoked and separated between tenants? - [Product / project](https://blog.openziti.io/ai-secops-why-your-ai-infrastructure-has-a-network-shaped-blind-spot) - [Documentation / policy](https://openziti.io/docs/learn/introduction/features/) ## Opper Category: Specialist gateways. Deployment: Managed. Reviewed: 2026-10-07. Security overview updated 5 October 2026: upstream inference is not EEA-restricted by default. Tracing, backups and provider retention have separate rules. Evaluation question: Does the selected route meet geography and retention needs, including backups? - [Product / project](https://opper.ai/) - [Documentation / policy](https://opper.ai/security-overview) ## orq.ai Category: Specialist gateways. Deployment: Managed, Customer-hosted. Reviewed: 2026-10-07. Gateway within an orchestration platform. Vendor security claims and logging retention must be matched to the purchased deployment. Evaluation question: Which attestations and retention settings cover this service and edition? - [Product / project](https://orq.ai/platform/ai-gateway) - [Documentation / policy](https://orq.ai/legal/security) ## Plano Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Agent proxy project formerly named Arch. Establish the current component boundaries and supported deployment before comparing controls. Evaluation question: Which tool, model and policy paths pass through Plano? - [Product / project](https://planoai.dev/) - [Documentation / policy](https://docs.planoai.dev/get_started/overview.html) ## Portkey / Prisma AIRS AI Gateway Category: Specialist gateways. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. Portkey commercial offerings have moved into Prisma AIRS. The selected advisory applies to the open-source Portkey gateway; it does not establish managed-service exposure. Evaluation question: Which product, build and custom-host policy are you evaluating? - [Product / project](https://portkey.ai/features/ai-gateway) - [Documentation / policy](https://www.paloaltonetworks.com/blog/2026/07/announcing-general-availability-of-prisma-airs-ai-gateway/) - [Related case](https://onequill.dev/resources/portkey-custom-host-ssrf) ## Requesty Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Gateway logging, caching and upstream retention are separate. EU router processing does not establish the upstream inference location. Evaluation question: Does this plan and route prohibit training and retention at every layer? - [Product / project](https://www.requesty.ai/) - [Documentation / policy](https://www.requesty.ai/privacy) ## SGLang Category: Inference engines. Deployment: Self-hosted. Reviewed: 2026-10-07. LLM and multimodal serving framework with distributed workers and prefix-cache reuse. The inference API, administrative controls and worker channels have distinct trust boundaries. Evaluation question: Which API and admin keys, worker networks, model-loading features and cache isolation are configured in the installed release? - [Project](https://github.com/sgl-project/sglang) - [Server arguments and access controls](https://docs.sglang.io/docs/advanced_features/server_arguments) - [Session-aware radix cache](https://github.com/sgl-project/sglang/blob/main/docs/docs/advanced_features/session_radix_cache.mdx) - [Related case](https://onequill.dev/resources/sglang-worker-and-management-trust) ## TensorZero Category: Specialist gateways. Deployment: Self-hosted. Reviewed: 2026-10-07. Gateway authentication is an explicit operational configuration with PostgreSQL support. Dashboard access is a separate boundary. Evaluation question: Are both the gateway and dashboard authenticated for this deployment? - [Product / project](https://www.tensorzero.com/) - [Documentation / policy](https://www.tensorzero.com/docs/operations/set-up-auth-for-tensorzero) ## Tetrate Agent Router Category: Specialist gateways. Deployment: Managed, Customer-hosted. Reviewed: 2026-10-07. Hosted Agent Router Service and enterprise customer-hosted data plane have different data flows. The management plane is separately hosted. Evaluation question: What crosses between your data plane and the hosted management plane? - [Product / project](https://www.tetrate.io/) - [Documentation / policy](https://docs.tetrate.ai/product-architecture/data-flows/) ## Traefik Hub Category: API platforms. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. Hub AI middlewares include Redis-backed shared token limits. Interrupted streams and response-completed events affect usage measurement. Evaluation question: Does incomplete-stream usage reach the budget ledger? - [Product / project](https://doc.traefik.io/traefik-hub/ai-gateway/overview) - [Documentation / policy](https://doc.traefik.io/traefik-hub/ai-gateway/middlewares/token-rate-limit) ## TrueFoundry Category: Specialist gateways. Deployment: Managed, Customer-hosted. Reviewed: 2026-10-07. Gateway, traces and deployment platform. Residency-aware fallback is a vendor-described configuration, not proof of your route policy. Evaluation question: Can the vendor show that all retries and fallback targets satisfy your residency policy? - [Product / project](https://www.truefoundry.com/ai-gateway) - [Documentation / policy](https://www.truefoundry.com/docs/ai-gateway/feedback-for-traces) ## Tyk AI Studio / MCP Gateway Category: API platforms. Deployment: Managed, Self-hosted. Reviewed: 2026-10-07. AI Studio, chat and MCP offerings are separate capabilities in the Tyk ecosystem. Evaluation question: Which model-routing and tool-access controls are available in the purchased product? - [Product / project](https://tyk.io/tyk-mcp-gateway/) - [Documentation / policy](https://tyk.io/blog/introducing-tyk-ai-studio-welcome-to-the-future-of-ai-governance-powered-by-ai-chat-gateway-and-portal/) ## Vercel AI Gateway Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Gateway content deletion and provider ZDR arrangements have separate scope and eligibility. Evaluation question: Which providers and plan-specific agreements implement the required ZDR setting? - [Product / project](https://vercel.com/ai-gateway) - [Documentation / policy](https://vercel.com/i/secure-ai-gateway) ## vLLM Category: Inference engines. Deployment: Self-hosted. Reviewed: 2026-10-07. Inference engine and serving stack. API authentication, internal networks, model loading and GPU scheduling are separate boundaries from a gateway. Evaluation question: Which engine, features, model format and cluster topology are deployed? - [Product / project](https://docs.vllm.ai/) - [Documentation / policy](https://docs.vllm.ai/en/latest/usage/security/) - [Related case](https://onequill.dev/resources/vllm-kv-transfer-network-isolation) - [Related case](https://onequill.dev/resources/vllm-model-loading-python-optimisation) - [Related case](https://onequill.dev/resources/vllm-multimodal-input-validation) - [Related case](https://onequill.dev/resources/vllm-gguf-gpu-memory-isolation) ## WSO2 API Manager Category: API platforms. Deployment: Self-hosted, Managed. Reviewed: 2026-10-07. AI token policies and backend throttling have different enforcement paths. Local per-node limits differ from distributed Traffic Manager policies. Evaluation question: Which counters are shared across replicas and regions? - [Product / project](https://apim.docs.wso2.com/en/latest/ai-gateway/rate-limiting/) - [Documentation / policy](https://apim.docs.wso2.com/en/latest/api-design-manage/design/rate-limiting/protect-backend-services/) ## ZenMux Category: Hosted routers. Deployment: Managed. Reviewed: 2026-10-07. Turning off Data Services changes ZenMux logging and related features. It does not independently establish upstream zero retention. Evaluation question: Which data services are disabled and what provider retention remains? - [Product / project](https://zenmux.ai/) - [Documentation / policy](https://zenmux.ai/docs/guide/advanced/data-services.html)