Cloud API management - verified
Azure API Management AI gateway
Azure API Management AI gateway is listed under cloud api management. Its first listed fee, in its own words: part of API Management (not a separate offering), billed with the API Management tier; capability availability varies by tier. It states OpenAI-compatible and Anthropic-compatible requests. Routing stated: fallbacks and load balancing. Read from the vendor's own pages on October 18, 2026.
Read from learn.microsoft.com. Docs - Compare every gateway
Status
Available on all API Management tiers; unified model API is in preview.
“unified model API (preview)”
learn.microsoft.com
- Status, as we read it
available on all API Management tiers; unified model API is in preview
“unified model API (preview)”
learn.microsoft.com
How Azure API Management AI gateway charges
Every fee line Azure API Management AI gateway states, in its own words and units, with the quote. Nothing is converted.
| Product | Fee, in the vendor's words | Applies to | Quote and source |
|---|---|---|---|
| Azure API Management AI gatewayPreview | part of API Management (not a separate offering) | billed with the API Management tier; capability availability varies by tier | “it's not a separate offering” |
Access and keys
- Free tier
- Not stated
Not stated on the checked page.
- Bring your own key
yes (you connect your own model deployments, including non-Microsoft providers)inferred
Our reading: Read as bring your own key: you connect your own model deployments, including other providers. The page does not use the term BYOK.
“Manage models deployed in Microsoft Foundry or providers such as Amazon Bedrock.”
learn.microsoft.com - BYOK fee
- Not stated
Not stated on the checked page.
API formats
- OpenAI-compatible
yes (OpenAI Chat Completions/Responses schemas; unified model API exposes an OpenAI-compatible endpoint, preview)
“It exposes multiple backends through a single OpenAI-compatible endpoint, handles format translation automatically, and lets you apply governance policies once across all models.”
learn.microsoft.com - Anthropic-compatible
yes (Anthropic Messages API, v2 tiers)
“Anthropic Messages API (currently supported in API Management v2 tiers)”
learn.microsoft.com - OpenAI base URL
- Not stated
Instance-specific; no fixed base URL.
- Anthropic base URL
- Not stated
Instance-specific; no fixed base URL.
Routing
- Fallbacks
circuit breaker and priority backendsinferred
Our reading: Read as fallback: the page describes circuit breaker rules that stop forwarding requests to a failing backend, plus priority backends. It does not name a fallback feature.
“You can also define circuit breaker rules to stop forwarding requests to AI service backends”
learn.microsoft.com - Load balancing
round-robin, weighted, priority-based, session-aware
“The backend load balancer supports round-robin, weighted, priority-based, and session-aware load balancing.”
learn.microsoft.com - Automatic model routing
- Not stated
Not stated on the checked page.
Caching
- Exact-match cache
- Not stated
Not stated on the checked page.
- Semantic cache
yes (Azure Managed Redis or RediSearch-compatible cache)
“In API Management, enable semantic caching by using Azure Managed Redis or another external cache compatible with RediSearch and onboarded to Azure API Management.”
learn.microsoft.com
Controls and observability
- Spend controls
token quotas per consumer
“you can set a TPM limit or a token quota over a specified period, such as hourly, daily, weekly, monthly, or yearly.”
learn.microsoft.com - Rate limits
token-per-minute limits
“Configure a token limit policy on your LLM APIs to manage and enforce limits per API consumer based on the usage of AI service tokens.”
learn.microsoft.com - Guardrails
Azure AI Content Safety moderation policy
“Apply policies to automatically moderate LLM prompts by using Azure AI Content Safety.”
learn.microsoft.com - Observability
prompt/completion logging to Azure Monitor; token metrics in Application Insights
“Log prompts and completions to Azure Monitor.”
learn.microsoft.com - Telemetry and log export
token metrics emitted to Azure Monitor
“Track token metrics per consumer in Application Insights.”
learn.microsoft.com
Data, deployment and license
- Data logging and retention
- Not stated
Not stated on the checked page.
- Regions
multi-region gateways (tier dependent)
“addition of regional gateways for multiregion deployments”
learn.microsoft.com - Self-hosting
manages self-hosted model endpointsinferred
Our reading: The quote is about managing self-hosted model endpoints, not about running the gateway itself on your own infrastructure.
“Self-hosted models and endpoints”
learn.microsoft.com - License
- Not stated
Commercial cloud service; no open-source license.
Vendor-stated counts
Model and provider counts as Azure API Management AI gateway states them, dated. When two pages disagree, both are shown.
No vendor-stated model or provider count on the checked page.
Research notes
The AI gateway is a set of capabilities inside Azure API Management, not a separate product; per-capability tier availability varies.
FAQ
How much does Azure API Management AI gateway cost?
In Azure API Management AI gateway's own terms: part of API Management (not a separate offering) (billed with the API Management tier; capability availability varies by tier). As of October 18, 2026; fees are not converted.
Does Azure API Management AI gateway support bring your own key?
Yes (you connect your own model deployments, including non-Microsoft providers) (our reading of the quote), per the vendor: "Manage models deployed in Microsoft Foundry or providers such as Amazon Bedrock.". BYOK fee: not stated. Not stated on the checked page.
Is Azure API Management AI gateway OpenAI-compatible or Anthropic-compatible?
OpenAI: yes (OpenAI Chat Completions/Responses schemas; unified model API exposes an OpenAI-compatible endpoint, preview). Anthropic: yes (Anthropic Messages API, v2 tiers).
Sources
Last verified from each vendor's own docs, pricing pages, READMEs and license files; every quote was string-matched against the page that day. Fees are shown in the vendor's own words and change without notice, so check the linked source before you commit.
Newsletter
Know when a gateway changes its fees
One email per week. New model releases, routing and pricing changes and what they mean for what you build.