Catatan
Akses ke halaman ini memerlukan otorisasi. Anda dapat mencoba masuk atau mengubah direktori.
Akses ke halaman ini memerlukan otorisasi. Anda dapat mencoba mengubah direktori.
Important
Azure Intelligent Recommendations Retirement and Migration Plan
Azure Intelligent Recommendations will be retired on March 31, 2026.
We're retiring the product due to changes in strategy and focus with the onset of the copilot era. Currently, there are no equivalent services available within the Microsoft stack, so we can't offer a migration plan to a similar service.
Next steps:
- Identify or build an alternative.
- Migrate all your traffic to the chosen alternative.
- Stop calling the Recommendations endpoint before the shutdown date.
- Delete all service endpoints, modeling, and IR accounts in the Azure portal before the shutdown date.
We'll continue to monitor and maintain Azure Intelligent Recommendations during the retirement period. After the product is retired, we'll keep the service data for three months before deleting it.
Pricing at a glance
Meet your organization’s recommendations needs and budget with competitive, pay-as-you-go pricing for creating AI models and serving recommendations results at scale.
Serving
You're charged for the real-time recommendation requests (transactions) processed per region by the Intelligent Recommendations service for each Service Endpoint, for each hour, counting the transactions in units of TPS-Hour, and using the tiered pricing table.
Standard and Premium requests are measured separately (see appendix for details on Standard and Premium features).
| TPS-Hour | Standard Price per TPS-Hour | Premium Price per TPS-Hour |
|---|---|---|
| Up to 20,000 per region | $0.20 | $0.40 |
| Next 180,000 per region | $0.10 | $0.30 |
| Above to 200,000 per region | $0.05 | $0.20 |
During account provisioning, you're asked to configure the “Preallocated Capacity” for each Service Endpoint for optimal working Transactions per seconds (TPS) to ensure proper SLAs, and the Service Endpoint can handle the allocated number of transactions per second.
The calculation of number of transactions handled by the Service Endpoint instance within an hour takes the MAX of (actual number of requests received, and the Preallocated Capacity) transactions every second and sums this number over the entire hour.
If you use multiple Modeling instances in the same Intelligent Recommendations account, then all recommendations requests are processed as Premium transactions.
Modeling
You're charged for the modeling compute hours consumed to train the AI-ML models.
The exact number of modeling hours consumed depends on the size of your data and catalog, the type of features and lists you model, and the frequency at which you model them.
There are two types of Modeling – Standard and Premium.
Modeling hours for Standard and Premium are counted separately and are charged differently (see appendix for details on Standard and Premium features).
Standard Modeling / Hour Price = $0.24
Premium Modeling / Hour Price = $0.50
There's a minimum charge of two (2) modeling hours a day per Modeling Instance.
Once modeling resources are inactive or deleted, you'll no longer be charged.
Free Trial
You can provision the Intelligent Recommendations service with one (1) Modeling instance and one (1) Service Endpoint instance with Preallocated Capacity set to one (1) TPS.
You can use the service in Trial mode for free for three (3) months. Trial period also ends if you provision more than one Modeling or Service Endpoint instances, or if you use Service Endpoint instances with Preallocated Capacity set to more than one (1) TPS.
Premium and Standard Features
Premium features
- Similar Looks
- Bring the Look
- Similar Text (NLP)
- Multiple Modeling instances on one (1) Intelligent Recommendations account
Standard features
- Personalized Picks for You
- People who do this tend to also do this
- Next Best Action
- Frequently Bought Together
- Real-time Session based Picks for You
- New
- Popular
- Trending
Pricing Example
This example is of a business using Intelligent Recommendations to suggest relevant items for a user in various different experiences. You can assume that the business uses a single Standard Intelligent Recommendations account and model, which is by default and automatically trained once per day.
This table shows how the monthly costs according to various Pre-Allocated Capacity values set for Service Endpoints, for all requests of Standard type. Costs below are in USD and include an assumed modeling cost of $300.
| Pre-Allocated Capacity for Service Endpoint | Total Monthly Cost |
|---|---|
| 5 | $1,020 |
| 10 | $1,420 |
| 15 | $1,780 |
| 20 | $2,140 |
| 30 | $2,860 |
| 50 | $4,300 |
| 100 | $6,300 |
| 200 | $9,900 |
| 500 | $20,700 |