Share via

Moving from Consumption Plan function app to Premium Plan

Chris Gray 10 Reputation points
2026-04-27T10:11:16.21+00:00

Good Morning,

I am looking to move from consumption plan to premium plan function apps. However, I am wanting to calculate the cost of what we will need in regards to a premium plan app. How do I work out what premium plan I will need (and number of instances) based on consumption plan? From what I can see there is no correlation between the two. Is there a conversion or a formula I can apply for this?

Many Thanks,

Chris

Azure Functions
Azure Functions

An Azure service that provides an event-driven serverless compute platform.

0 comments No comments

2 answers

Sort by: Most helpful
  1. Siddhesh Desai 6,635 Reputation points Microsoft External Staff Moderator
    2026-04-27T14:35:08.1066667+00:00

    Hi @Chris Gray

    Thank you for reaching out to Microsoft Q&A.

    When moving from an Azure Functions Consumption plan to a Premium (Elastic Premium) plan, there is no direct conversion or formula that can be applied. This is because the two plans use completely different billing and scaling models. The Consumption plan is billed based on executions and GB‑seconds and scales automatically without any pre‑provisioned instances, whereas the Premium plan is billed based on pre‑allocated compute (vCPU and memory) with one or more always‑on instances. Due to this architectural difference, Microsoft does not provide a 1:1 mapping or automated calculator to convert Consumption usage directly into a Premium plan size.

    Refer below points to resolve this issue or as a workaround:

    Analyze current Consumption plan usage

    Review Azure Monitor and Application Insights metrics such as average and peak execution duration, GB‑seconds consumed, concurrency, and trigger patterns. These metrics indicate how much compute your functions actually use over time and are the foundation for Premium sizing.

    Estimate required always‑on compute for Premium

    Translate sustained GB‑seconds usage into continuously allocated memory and CPU. For example, steady usage close to 1 GB‑second per second generally maps to a single EP1 instance. Short‑lived spikes can be handled by Premium auto‑scale, while sustained peaks require more instances or a larger SKU.

    Select an initial Premium SKU and instance count

    Start with a conservative configuration such as EP1 with minimum instances set to 1, then monitor CPU, memory, and scale‑out behavior. Increase the SKU (EP2/EP3) or maximum instances only if the workload consistently demands it.

    Validate cost using Azure Pricing Calculator and monitoring Use observed metrics (instance size, instance count, and running hours) in the Azure Pricing Calculator to estimate monthly cost. After migration, continuously monitor usage and fine‑tune the plan to balance performance and cost.


  2. Divyesh Govaerdhanan 10,885 Reputation points MVP Volunteer Moderator
    2026-04-27T14:33:32.1166667+00:00

    Hi Chris Gray,

    Welcome to Microsoft Q&A,You're right that there's no direct conversion formula between Consumption and Premium billing, because the two models charge for fundamentally different things.The Consumption plan bills based on observed resource consumption measured in gigabyte-seconds: average memory size in GB multiplied by execution time in milliseconds.

    The Premium plan bills based on the number of core seconds and memory allocated across instances, with no per-execution charge, and at least one instance is always active and billed whether your functions are running or idle.

    You're moving from paying per execution to paying for reserved compute, so the right question isn't "how much did I consume?" but "how much compute do I need allocated at any point in time?"

    Before choosing a SKU, go to your existing Consumption plan function app in Azure Monitor and look at these metrics over the last 30 days:

    • Function Execution Units (GB-ms): this tells you total GB-seconds consumed and helps you understand memory pressure across all executions
    • Function Execution Count: total invocations, which helps you understand concurrency
    • Max instance count reached: check the scale controller logs or Application Insights to understand peak concurrent instances the Consumption plan spun up

    Premium plan workers are available in three fixed sizes: EP1 with 1 vCPU and 3.5 GB RAM, EP2 with 2 vCPU and 7 GB RAM, and EP3 with 4 vCPU and 14 GB RAM.

    For most workloads migrating from Consumption, EP1 with 1 minimum instance and a burst cap of 3 to 5 is a reasonable starting point. The Premium plan will auto-scale using the same event-driven logic as Consumption, so you only pay for the burst instances while they're active.

    Run the numbers through the Azure Pricing Calculator using your execution count and duration data. The Premium plan will be more expensive at low traffic volumes (because of the always-on minimum instance), but typically cheaper at sustained high throughput where Consumption GB-s charges add up.

    Please Upvote and accept the answer if it helps!!


Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.