IBM Quantum Plan: Choose Platform & Access

To use IBM Quantum, you choose a plan (Open, Pay-As-You-Go, or Premium) that determines which quantum computing platform instances you can access, and then you connect to those instances through the IBM Quantum Platform console or via API. The plan you select directly controls the number of qubits, the quantum processors available, and the execution limits you get. This article explains the tiers, how to select an instance, and what to consider when choosing.

Background: IBM Quantum Plans and Instances

IBM Quantum offers cloud access to its family of superconducting quantum processors. The access is organized into plans, each with different quotas and available hardware. An instance is a specific combination of a quantum processor (like a specific Eagle or Heron chip) and a set of access parameters (such as monthly execution seconds and queue priority). When you submit a job, you target it at an instance, not just at a generic “quantum computer.”

Historically, IBM offered a free “Open” plan and a paid “Premium” plan, but in 2024 they introduced a more granular structure. The current main plans are:

Open Plan (Free): Access to a limited set of processors, typically with lower qubit counts and fewer execution seconds per month. It’s designed for learning, testing, and small-scale research.
Pay-As-You-Go (PAYG): A flexible plan where you pay for the compute time you use. It offers access to more powerful processors (like 127-qubit Eagle and 133-qubit Heron) without a monthly subscription. You are billed per second of quantum time.
Premium Plan: A subscription-based plan for organizations that need dedicated capacity, higher quotas, and priority queue. It includes access to the full fleet of IBM Quantum processors, including the most advanced ones.

Each plan has a set of instances you can choose from. For example, on the Open plan you might see instances like ibm_brisbane or ibm_kyiv, while Premium users see the same instances but with higher monthly quotas and faster queue times.

How to Choose a Plan and Access an Instance

Choosing the right plan depends on your goals, budget, and required performance. Here’s a practical guide:

Assess your needs: Are you a student exploring quantum computing, a researcher running experiments, or a business developing quantum applications? For learning, the Open plan is sufficient. For serious research or development, consider PAYG or Premium.
Check processor availability: The IBM Quantum Platform dashboard shows which processors are online and which plans they belong to. Some processors are reserved for Premium users, so if you need a specific chip, verify it’s in your plan.
Review quotas: Each instance has a monthly execution time limit (e.g., 10 minutes on Open, 1000 minutes on PAYG). If you exceed it, you must wait until the next month or upgrade.
Consider queue priority: On the Open plan, jobs often sit in a queue for hours. PAYG and Premium give you higher priority, meaning faster turnaround.
Sign up and authenticate: Go to quantum.ibm.com, create an IBMid, and log in. The platform automatically assigns you to the Open plan. To upgrade, you need to provide billing details.
Select an instance in code: When using Qiskit, you specify the instance as service = QiskitRuntimeService(instance="ibm-q/open/main") or a specific one like ibm-q-community/ibm-q/open/main. The exact string is shown in the dashboard under each processor.

Understanding the IBM Quantum Platform Console

The IBM Quantum Platform (quantum.ibm.com) is your central hub. After logging in, you see a dashboard with:

Compute: A list of available processors, their status (online/offline), and the plan they belong to.
Jobs: A history of your submitted jobs, their status, and results.
Instances: Your current instances, showing the plan, quota usage, and the processors accessible.
Documentation: Tutorials, API references, and Qiskit guides.

To access an instance programmatically, you use an API token from your account settings. This token is used in Qiskit to authenticate. For example:

from qiskit_ibm_runtime import QiskitRuntimeService
service = QiskitRuntimeService(channel="ibm_quantum", token="YOUR_TOKEN", instance="ibm-q-community/ibm-q/open/main")

Common Problems and Misconceptions

One frequent mistake is confusing plan with instance. The plan is your subscription tier; the instance is the specific processor+quota combination you use. You cannot choose an instance that is not part of your plan. For example, on the Open plan you cannot access ibm_sherbrooke if it’s reserved for Premium.

Another issue is hitting quota limits. The Open plan gives you a small monthly allocation (often 10 minutes of quantum time). If you run many jobs, you quickly exhaust it. The error message “Quota exceeded” means you must wait or upgrade.

Also, not all processors are available 24/7. IBM periodically brings down systems for calibration. Check the status before submitting critical jobs.

Costs and Pricing Structure

The Open plan is free, but with limited resources. The Pay-As-You-Go plan charges per second of quantum time, with different rates per processor (e.g., $1.60 per second for 127-qubit processors). The Premium plan is a custom-priced annual subscription. There are no hidden fees for classical simulation or data storage, but be aware that complex jobs may require more quantum time.

For budgeting, estimate your monthly quantum time. A typical experiment might use 10-100 seconds. If you run experiments daily, PAYG could cost a few hundred dollars per month. Premium is cost-effective for organizations with heavy usage.

Comparing IBM Quantum Plans

Feature	Open	Pay-As-You-Go	Premium
Cost	Free	Pay per second	Annual subscription
Processor access	Limited (e.g., 7-qubit, 27-qubit)	Most processors	All processors
Monthly quota	~10 minutes	Varies, typically 1000 minutes	Custom, high
Queue priority	Low	Medium	High
Support	Community	Standard	Dedicated

Alternatives to IBM Quantum

If IBM Quantum doesn’t fit your needs, other cloud quantum providers include:

Amazon Braket: Access to multiple quantum hardware providers (IonQ, Rigetti, etc.) with a pay-per-use model.
Microsoft Azure Quantum: Similar multi-vendor access, with integration into Azure.
Google Quantum AI: Offers access to Sycamore processor via their Quantum Computing Service.
IonQ: Direct access to trapped-ion quantum computers.

Each has different hardware architectures, so the choice depends on the algorithm you want to run.

Frequently Asked Questions

Can I switch plans later?

Yes, you can upgrade or downgrade at any time. Upgrading to PAYG or Premium is immediate, but downgrading might take effect at the next billing cycle. Your existing jobs will finish, but you may lose access to certain instances.

Do I need to install Qiskit to use IBM Quantum?

Qiskit is the recommended SDK, but you can also use the web interface to submit jobs without installing anything. For programmatic access, you need Qiskit or another library that supports IBM Quantum’s API.

What is the difference between a backend and an instance?

A backend is the physical quantum processor (e.g., ibm_brisbane). An instance is a combination of a backend, a plan, and a project. In Qiskit, you specify the instance, and it includes the backend.

How do I see my current quota usage?

In the IBM Quantum Platform dashboard, go to the “Instances” section. Each instance shows a progress bar of your monthly usage. You can also check via API using service.instances().

Can I use IBM Quantum for free forever?

Yes, the Open plan is free indefinitely, but with limited resources. It’s suitable for learning and small experiments. For larger workloads, you’ll need a paid plan.

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.

More cards in this drawer

The whole Numbers and Powers drawer →