Choose OmniSwitch when…
- You want a compact control plane without adopting a Kubernetes-native gateway layer.
- A binary, Docker, or a small self-hosted deployment meets your needs.
- You value local policy and explicit protocol boundaries.
Compare a secure self-hosted control plane, AI gateway routing, and agent protocol scope with Agent Router. Evaluate the operating model before the feature list.
Updated · Public beta
Formerly Envoy AI Gateway. Register a8a594b; each cell records its own version and verified date. A capability review, not a benchmark.
No scores or totals: uneven documentation coverage is not a measure of product quality.
| Capability | OmniSwitch[1]Register a8a594b | Agent Router[7]Formerly Envoy AI Gateway |
|---|---|---|
| Deployment & license | ||
| Self-hosted deployment | Not verifiedNo claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ | Not verifiedNo claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| SQLite / Postgres storage | SupportedSQLite by default; PostgreSQL via OMNISWITCH_DATABASE_URL. Logs and keys remain SQLite-local even in the Postgres case per the project's own docs.
| Not availableState is Kubernetes CRDs / Envoy xDS config, not an embedded SQL database.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Kubernetes manifests / Helm chart | SupportedKustomize manifests (Deployment/Service/ConfigMap/Secret example/PVCs/Redis) plus a Helm chart; no operator/CRDs/inference-aware routing.
| SupportedNative Kubernetes Gateway API controller with an official Helm chart; this is the project's only real deployment target (no practical loopback/single-binary mode for a fair same-host comparison).
|
| Config hot reload | SupportedConfig hot reload covers routes, guardrails, cache posture, circuit breaker, and shadow provider without restart.
| Not verifiedKubernetes CRD reconciliation implies config changes without a full restart, but this was not independently confirmed as "hot reload" in the project's own terms.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Built-in dashboard/UI | SupportedLocal dashboard for logs, metrics, prompts, providers, and guardrails.
| Not availableNo bundled dashboard found; this is a Kubernetes control-plane/CRD project, not a single binary with a UI.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| License | Supported | Not verifiedThe former envoyproxy/ai-gateway GitHub repository now redirects to theagentrouter/agent-router (project renamed to "Agent Router" during 2026; the project states images, CRDs, CLI (aigw), and Go module path are unchanged). License metadata read back from the redirect target (Apache-2.0) but this may reflect a different/renamed project's repo settings rather than a re-verified license for the exact codebase; treat as unknown pending direct confirmation.
|
| APIs & providers | ||
| OpenAI-compatible chat completions | Supported/v1/chat/completions, streaming and non-streaming.
| SupportedOpenAI-compatible routing documented.
|
| OpenAI-compatible embeddings | Supported | Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| OpenAI Responses API | Partial/v1/responses text subset only (no image/audio/tool-streaming parity with OpenAI).
/v1/responses text subset only (no image/audio/tool-streaming parity with OpenAI). | Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Anthropic Messages API | Partial/v1/messages core subset. | Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Native (non-passthrough) provider adapters | Partial4 native adapters (OpenAI, Anthropic, Google, Groq); no native Bedrock/Vertex/Cohere. Everything else goes through the generic OpenAI-compatible custom adapter with 15+ documented presets.
4 native adapters (OpenAI, Anthropic, Google, Groq); no native Bedrock/Vertex/Cohere. Everything else goes through the generic OpenAI-compatible custom adapter with 15+ documented presets. | SupportedModel virtualization across multiple providers documented.
|
| Custom/generic OpenAI-compatible endpoint | SupportedGeneric custom provider adapter with presets for Azure OpenAI, Ollama, vLLM, LiteLLM, DeepSeek, Together, Fireworks, and aliases.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Routing, reliability & cost | ||
| Routing & fallbacks | Supportedroute-level fallbacks, max_retries, retry_backoff, retry_codes, per-attempt timeout.
| SupportedProvider fallback / automatic failover documented.
|
| Weighted / load-balanced routing | SupportedWeighted variants per route.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Conditional routing (rule/expression based) | SupportedCEL condition expressions on route variants.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Circuit breaker | SupportedDefault 5 failures / 60s cooldown, configurable via gateway.circuit_breaker_failures / _cooldown.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Exact-match cache | SupportedSHA-256 request-fingerprint cache key, tenant-scoped (api_key/workspace/organization/global), SQLite-backed.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Similarity cache (see scope) | SupportedNOT embedding-based. Vectorize() builds a normalized token-frequency map over lowercased, stop-word-filtered tokens; Similarity() is cosine over those token-frequency vectors. State this precisely wherever OmniSwitch's cache is compared to embedding-based semantic caches.
| Not availableThis project's own homepage previously claimed 'no built-in semantic caching' for this gateway ("Semantic Caching" false in the earlier comparison), matching the absence of any caching claim in this gateway's own capability docs.
|
| Budgets and quotas | SupportedCost and token budgets per API key.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Rate limiting | PartialLocal sliding-window request limits plus optional Redis-coordinated fixed window across replicas; no token-based (TPM) limiter.
Local sliding-window request limits plus optional Redis-coordinated fixed window across replicas; no token-based (TPM) limiter. | SupportedUsage-based, token-aware rate limiting and quota-aware routing documented.
|
| Token-based rate limits | Not verifiedNo claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ | Not verifiedNo claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Identity & guardrails | ||
| API key management | SupportedHashed API keys, bootstrap owner key, workspace-scoped keys.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| OIDC / JWT identity | SupportedOIDC JWT/JWKS workload identity (identity.oidc.jwks_url/issuer/audience).
| Not verifiedOAuth authentication is documented for the MCP gateway feature specifically; general gateway-level OIDC/JWT auth not independently confirmed.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Mutual TLS | Not availableExplicitly not implemented; listed as a roadmap item, not a shipped feature.
| Not verifiedEnvoy proxies commonly support mTLS termination, but this was not confirmed as a documented capability of this specific project.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| RBAC / policy engine (CEL or equivalent) | SupportedCEL allow/deny authorization rules plus fixed role gates.
| Not verifiedEnvoy Gateway's underlying policy model supports xDS/RBAC filters; a CEL-specific policy surface for this AI gateway project was not confirmed.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Built-in guardrails | SupportedBuilt-in PII/prompt-injection/SQL-injection/toxicity/secret detectors plus custom regex rules, single-pass Rust-WASM scan with pure-Go fallback.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Guardrail webhooks / external connectors | SupportedHTTP webhook connectors with configurable stage/action/timeout/fail_open.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Output redaction | Supportedredact action available on output-stage guardrail rules.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Guardrails applied to streaming output | Supportedguardrails.stream_buffer defaults to true: the full SSE stream is buffered and scanned before release, which means TTFT with guardrails on is dominated by upstream generation time, not gateway overhead. Benchmark both stream_buffer=true and false and label them.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Governed agent infrastructure | ||
| MCP gateway | SupportedHTTP MCP targets with streamed SSE / streamable-HTTP pass-through.
| SupportedFull MCP gateway with server multiplexing, tool routing/filtering, OAuth, and observability documented.
|
| MCP over stdio | SupportedPersistent stdio MCP transport targets.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| MCP federation (multiple targets) | SupportedMultiple named MCP targets, policy-gated tools/call, server-side headers.
| SupportedMCP server multiplexing across multiple targets documented.
|
| OpenAPI-to-MCP conversion | Not availableNot implemented; explicit gap in the project's own comparison doc.
| SupportedThis project's internal comparison table lists this gateway (as agentgateway's sibling ecosystem entry / envoy-ai-gateway) as "Yes" for OpenAPI-to-MCP conversion; not independently re-verified against this project's current docs.
|
| MCP OAuth / token exchange | SupportedRFC 8693 token exchange plus RFC 7523 JWT Bearer Grant for MCP OAuth delegation.
| SupportedOAuth authentication for the MCP gateway documented.
|
| A2A agent card discovery | SupportedPublic A2A v1 Agent Card discovery (/.well-known/agent-card.json) plus authenticated GetExtendedAgentCard.
Current protocol scope: SendMessage, SendMessageStream (SSE task events), GetTask, best-effort CancelTask; tasks in memory; no push notifications; no downstream agent proxy. | Not availableNo evidence of native A2A support found; the documented agent/tool integration focus is MCP.
|
| A2A task methods | PartialSendMessage / SendMessageStream, GetTask, CancelTask implemented (JSON-RPC methods and REST aliases). Agent card advertises pushNotifications: false, so push-notification task updates are not supported.
Current protocol scope: SendMessage, SendMessageStream (SSE task events), GetTask, best-effort CancelTask; tasks in memory; no push notifications; no downstream agent proxy. SendMessage / SendMessageStream, GetTask, CancelTask implemented (JSON-RPC methods and REST aliases). Agent card advertises pushNotifications: false, so push-notification task updates are not supported. | Not availableSame as above.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Operations & evidence | ||
| Prompt management | PartialTemplates with version history and rendering; no approval workflow, rollback/promotion, or experiments.
Templates with version history and rendering; no approval workflow, rollback/promotion, or experiments. | Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Evals / experiments | PartialPolicy replay endpoint for batch allow/deny simulation only; no model-quality eval datasets/scoring.
Policy replay endpoint for batch allow/deny simulation only; no model-quality eval datasets/scoring. | Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| OTLP telemetry export | SupportedOTel trace export (tested against Langfuse/Jaeger/Tempo).
| SupportedEnterprise observability documented; Envoy's ecosystem is commonly OTel-instrumented (e.g. Apache SkyWalking's Envoy AI Gateway monitoring guide).
|
| Prometheus metrics | Supported/metrics endpoint, gated by OMNISWITCH_PROMETHEUS_ENABLED.
| Not verifiedNot confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
| Published cross-gateway benchmarks | PartialIn-repo Rust-WASM-vs-Go guardrail micro-benchmark only (BENCHMARKS.md, measured on a non-pinned Windows laptop). No cross-gateway comparative run published yet -- this benchmark effort (bench/) is that follow-up.
In-repo Rust-WASM-vs-Go guardrail micro-benchmark only (BENCHMARKS.md, measured on a non-pinned Windows laptop). No cross-gateway comparative run published yet -- this benchmark effort (bench/) is that follow-up. | Not availableNo published cross-gateway overhead benchmark found for this project specifically.
No capability-specific source was recorded. The register is the provenance, not proof of absence. Open pinned register ↗ |
No claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗No claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗SQLite by default; PostgreSQL via OMNISWITCH_DATABASE_URL. Logs and keys remain SQLite-local even in the Postgres case per the project's own docs.
State is Kubernetes CRDs / Envoy xDS config, not an embedded SQL database.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Kustomize manifests (Deployment/Service/ConfigMap/Secret example/PVCs/Redis) plus a Helm chart; no operator/CRDs/inference-aware routing.
Native Kubernetes Gateway API controller with an official Helm chart; this is the project's only real deployment target (no practical loopback/single-binary mode for a fair same-host comparison).
Config hot reload covers routes, guardrails, cache posture, circuit breaker, and shadow provider without restart.
Kubernetes CRD reconciliation implies config changes without a full restart, but this was not independently confirmed as "hot reload" in the project's own terms.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Local dashboard for logs, metrics, prompts, providers, and guardrails.
No bundled dashboard found; this is a Kubernetes control-plane/CRD project, not a single binary with a UI.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗The former envoyproxy/ai-gateway GitHub repository now redirects to theagentrouter/agent-router (project renamed to "Agent Router" during 2026; the project states images, CRDs, CLI (aigw), and Go module path are unchanged). License metadata read back from the redirect target (Apache-2.0) but this may reflect a different/renamed project's repo settings rather than a re-verified license for the exact codebase; treat as unknown pending direct confirmation.
/v1/chat/completions, streaming and non-streaming.
OpenAI-compatible routing documented.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗/v1/responses text subset only (no image/audio/tool-streaming parity with OpenAI).
/v1/responses text subset only (no image/audio/tool-streaming parity with OpenAI).
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗/v1/messages core subset.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗4 native adapters (OpenAI, Anthropic, Google, Groq); no native Bedrock/Vertex/Cohere. Everything else goes through the generic OpenAI-compatible custom adapter with 15+ documented presets.
4 native adapters (OpenAI, Anthropic, Google, Groq); no native Bedrock/Vertex/Cohere. Everything else goes through the generic OpenAI-compatible custom adapter with 15+ documented presets.
Model virtualization across multiple providers documented.
Generic custom provider adapter with presets for Azure OpenAI, Ollama, vLLM, LiteLLM, DeepSeek, Together, Fireworks, and aliases.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗route-level fallbacks, max_retries, retry_backoff, retry_codes, per-attempt timeout.
Provider fallback / automatic failover documented.
Weighted variants per route.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗CEL condition expressions on route variants.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Default 5 failures / 60s cooldown, configurable via gateway.circuit_breaker_failures / _cooldown.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗SHA-256 request-fingerprint cache key, tenant-scoped (api_key/workspace/organization/global), SQLite-backed.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗NOT embedding-based. Vectorize() builds a normalized token-frequency map over lowercased, stop-word-filtered tokens; Similarity() is cosine over those token-frequency vectors. State this precisely wherever OmniSwitch's cache is compared to embedding-based semantic caches.
This project's own homepage previously claimed 'no built-in semantic caching' for this gateway ("Semantic Caching" false in the earlier comparison), matching the absence of any caching claim in this gateway's own capability docs.
Cost and token budgets per API key.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Local sliding-window request limits plus optional Redis-coordinated fixed window across replicas; no token-based (TPM) limiter.
Local sliding-window request limits plus optional Redis-coordinated fixed window across replicas; no token-based (TPM) limiter.
Usage-based, token-aware rate limiting and quota-aware routing documented.
No claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗No claim is recorded for this capability in the pinned register.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Hashed API keys, bootstrap owner key, workspace-scoped keys.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗OIDC JWT/JWKS workload identity (identity.oidc.jwks_url/issuer/audience).
OAuth authentication is documented for the MCP gateway feature specifically; general gateway-level OIDC/JWT auth not independently confirmed.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Explicitly not implemented; listed as a roadmap item, not a shipped feature.
Envoy proxies commonly support mTLS termination, but this was not confirmed as a documented capability of this specific project.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗CEL allow/deny authorization rules plus fixed role gates.
Envoy Gateway's underlying policy model supports xDS/RBAC filters; a CEL-specific policy surface for this AI gateway project was not confirmed.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Built-in PII/prompt-injection/SQL-injection/toxicity/secret detectors plus custom regex rules, single-pass Rust-WASM scan with pure-Go fallback.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗HTTP webhook connectors with configurable stage/action/timeout/fail_open.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗redact action available on output-stage guardrail rules.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗guardrails.stream_buffer defaults to true: the full SSE stream is buffered and scanned before release, which means TTFT with guardrails on is dominated by upstream generation time, not gateway overhead. Benchmark both stream_buffer=true and false and label them.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗HTTP MCP targets with streamed SSE / streamable-HTTP pass-through.
Full MCP gateway with server multiplexing, tool routing/filtering, OAuth, and observability documented.
Persistent stdio MCP transport targets.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Multiple named MCP targets, policy-gated tools/call, server-side headers.
MCP server multiplexing across multiple targets documented.
Not implemented; explicit gap in the project's own comparison doc.
This project's internal comparison table lists this gateway (as agentgateway's sibling ecosystem entry / envoy-ai-gateway) as "Yes" for OpenAPI-to-MCP conversion; not independently re-verified against this project's current docs.
RFC 8693 token exchange plus RFC 7523 JWT Bearer Grant for MCP OAuth delegation.
OAuth authentication for the MCP gateway documented.
Public A2A v1 Agent Card discovery (/.well-known/agent-card.json) plus authenticated GetExtendedAgentCard.
Current protocol scope: SendMessage, SendMessageStream (SSE task events), GetTask, best-effort CancelTask; tasks in memory; no push notifications; no downstream agent proxy.
No evidence of native A2A support found; the documented agent/tool integration focus is MCP.
SendMessage / SendMessageStream, GetTask, CancelTask implemented (JSON-RPC methods and REST aliases). Agent card advertises pushNotifications: false, so push-notification task updates are not supported.
Current protocol scope: SendMessage, SendMessageStream (SSE task events), GetTask, best-effort CancelTask; tasks in memory; no push notifications; no downstream agent proxy.
SendMessage / SendMessageStream, GetTask, CancelTask implemented (JSON-RPC methods and REST aliases). Agent card advertises pushNotifications: false, so push-notification task updates are not supported.
Same as above.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Templates with version history and rendering; no approval workflow, rollback/promotion, or experiments.
Templates with version history and rendering; no approval workflow, rollback/promotion, or experiments.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗Policy replay endpoint for batch allow/deny simulation only; no model-quality eval datasets/scoring.
Policy replay endpoint for batch allow/deny simulation only; no model-quality eval datasets/scoring.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗OTel trace export (tested against Langfuse/Jaeger/Tempo).
Enterprise observability documented; Envoy's ecosystem is commonly OTel-instrumented (e.g. Apache SkyWalking's Envoy AI Gateway monitoring guide).
/metrics endpoint, gated by OMNISWITCH_PROMETHEUS_ENABLED.
Not confirmed in available sources.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗In-repo Rust-WASM-vs-Go guardrail micro-benchmark only (BENCHMARKS.md, measured on a non-pinned Windows laptop). No cross-gateway comparative run published yet -- this benchmark effort (bench/) is that follow-up.
In-repo Rust-WASM-vs-Go guardrail micro-benchmark only (BENCHMARKS.md, measured on a non-pinned Windows laptop). No cross-gateway comparative run published yet -- this benchmark effort (bench/) is that follow-up.
No published cross-gateway overhead benchmark found for this project specifically.
No capability-specific source was recorded. The register is the provenance, not proof of absence.
Open pinned register ↗OmniSwitch scope. The pinned register marks OpenAI Responses API, Anthropic Messages API, Native (non-passthrough) provider adapters, Rate limiting, Mutual TLS, OpenAPI-to-MCP conversion, A2A task methods, Prompt management, Evals / experiments, Published cross-gateway benchmarks as partial or unavailable. Read each source and version; compare them with the build you actually deploy.
Read unknowns carefully. “Not verified” means an evidence gap, not absence. Editions remain separate. Register a8a594b; 2 compared deployments.
A2A (Beta). SendMessage, SendMessageStream (SSE task events), GetTask, best-effort CancelTask; tasks in memory; no push notifications; no downstream agent proxy.
Agent Router has a different platform or proxy operating model. Start a separate OmniSwitch deployment, define a representative route, and compare request behavior and operational fit. No base-URL migration guide is offered without verified compatibility on both sides.
Start a local evaluationThe project was renamed Agent Router in v1.2.0 on 6 October 2026. Existing API groups, CRDs, charts, images, and CLI names remain unchanged. It is not the separate AgentGateway project.
Source · accessed 2026-10-09Product register a8a594b. Cells reflect registered source claims, not roadmap plans or release-artifact availability; each retains its own verified date and version.
Formerly Envoy AI Gateway. Capability-specific documentation and dated editorial sources remain separate from independent testing.
Supported means documented within the named capability and deployment scope, not full product parity. Beta and licensed editions are shown when the register supplies a qualifier. Not verified describes an evidence gap, not absence. Claims come from the pinned register; cells have no scores or totals. This is a dated documentation review, not a performance test or a security certification. See the benchmark methodology.