Anaplan's AI stack has expanded fast, and the terminology has expanded with it. CoModeler, Agent Studio, Forecaster, Hyperblock, xP&A, Analyst agents — they seem related, and indeed they are, but each does its own thing in Connected Planning. This glossary explains exactly what each term means, how they integrate, and how planning executives should think about them when evaluating their options.
Polestar Analytics is a preferred global AI delivery partner and global implementation partner for CoModeler and Agent Studio by Anaplan, and hence, the view presented in this guide will be from that perspective.
Behind the answer above, the stack breaks down cleanly across four working layers, each owning a distinct job:
- Hyperblock does the math.
- CoModeler builds and refines the model.
- Forecaster runs the prediction.
- Agent Studio governs who can do what.
All of it inherits your existing workspace permissions, so nothing operates outside your security perimeter.
Anaplan CoModeler is the Anaplan AI model builder. You describe what you need in plain English or hand it a CSV spec, and it drafts the lists, modules, line items, and formulas, all aligned to DISCO methodology and the Planual. Nothing gets applied until a human approves it.
Core capabilities:
- Generates lists, modules, line items, and formulas from natural language prompts or CSV specs
- Extends existing Anaplan models with new logic, dimensions, or scenarios without manual dependency tracing
- Reviews live models for inefficiencies, redundant formulas, and error-prone patterns
- Produces step-by-step build plans for reviewer approval before any change is applied
- Works across both Classic and Polaris calculation engines
- Grounded in DISCO methodology and the Planual, so outputs follow Anaplan best practices by default
Anaplan Agent Studio is the control room for every Anaplan AI agent in your environment: custom analysts, role-based agents, and CoModeler itself. It is where an AI admin decides which agents exist, what data they can see, and which workspaces they run in, all under existing role and dimension security. For enterprises structuring rollout without governance drift, this practical setup guide is worth reviewing.
Core capabilities:
- Centralized admin console to build, configure, test, and deploy all Anaplan AI agents
- Manages custom analysts, role-based agents, and CoModeler from one place
- Enforces workspace boundaries and inherits existing role and dimension security
- Data source configuration at the module and line-item level per agent
- Suggested starter questions to accelerate user adoption
- Sync logs and scheduled data refreshes for every agent
- Controlled activation of CoModeler at the workspace level for intentional rollouts
Anaplan Forecaster is the intelligent forecasting layer inside Anaplan. It applies statistical and machine learning models to demand, revenue, and workforce planning, so forecasts live inside the same model your teams are already planning against. Teams looking at how the earlier PlanIQ foundations connect to today's forecasting stack can start with this overview of AI-enabled forecasting in Anaplan.
Core capabilities:
- Statistical and ML-driven forecasts run natively inside planning models
- Supports demand, revenue, workforce, and financial forecasting use cases
- Handles seasonality, promotions, and external drivers as forecast inputs
- Outputs are reviewable and adjustable inside the same model planners already use
- No separate data science stack required, since it runs on Anaplan's own compute layer
Anaplan xP&A stands for Extended Planning & Analysis. It is what happens when finance stops planning in isolation and starts planning on the same model as sales, supply chain, and workforce. For senior leaders deciding which functions to unify first, this overview of Anaplan applications across the enterprise is a useful starting point.
Core capabilities:
- In-memory calculation with real-time recalculation across dependent cells
- Multi-dimensional modelling supporting concurrent enterprise-wide planning
- Foundation layer that every Anaplan AI feature runs on top of
- Supports both the Classic engine and the newer Polaris engine for sparse, high-dimensional models
Anaplan xP&A stands for Extended Planning & Analysis. It is what happens when finance stops planning in isolation and starts planning on the same model as sales, supply chain, and workforce.
Core capabilities:
- Shared planning models across finance, sales, supply chain, workforce, and operations
- Real-time cross-functional scenario planning on a single source of truth
- Delivered through Anaplan Applications and custom models running on Hyperblock
- Supported by role-based AI agents that surface function-specific insights
Anaplan AI agents are the conversational agents — Finance Analyst, Supply Chain Analyst, Sales Analyst, Workforce Analyst — that answer planning questions against your live model. The Anaplan AI assistant is CoModeler, sitting alongside model builders through the build, extend, and optimize lifecycle.
Core capabilities:
- Natural-language querying of live planning models
- Role-based agents pre-built for finance, supply chain, sales, and workforce
- Custom Analyst agents configurable for bespoke models through Agent Studio
- Inherit existing user permissions, so answers respect who is asking
- CoModeler acts as the AI assistant for model builders across build, extend, and optimize workflows
| Capability |
Purpose |
Used By |
| CoModeler |
Build and optimize models |
Model Builders |
| Agent Studio |
Govern AI agents |
AI Admins / COE |
| Forecaster |
Predict future outcomes |
Finance, Supply Chain |
| Hyperblock |
Execute calculations |
Entire platform |
| xP&A |
Connect business planning |
Executive Leadership |
| Analyst Agents |
Answer planning questions in natural language |
Function Leads |
Polestar Analytics is one of 12 preferred global AI delivery partners for Anaplan CoModeler and Agent Studio, with 175+ certified architects across APAC and the US. For enterprises deciding how to sequence CoModeler, Agent Studio, and Forecaster inside an existing Connected Planning environment, the Polestar Analytics Anaplan practice page covers the delivery approach, and its broader enterprise planning services show how xP&A gets operationalized across finance, sales, supply chain, and workforce.
Polestar Analytics is an Anaplan implementation partner that builds AI-enabled connected planning solutions for finance and supply chain teams.
- Treating CoModeler as auto-modeling. It is human-in-the-loop; builder approval is mandatory.
- Deploying custom agents outside Agent Studio, which creates governance drift the moment you scale past two or three agents.
- Confusing Hyperblock with a database. It is a live calculation engine, and modelling against it that way changes what you can build.
- Positioning xP&A as a finance-only initiative. If sales and supply chain are not on the same model, it is still siloed planning with a new label.
- Underestimating data quality and model design before switching on AI. Agents surface existing problems faster; they do not fix them.
CoModeler greatly speeds up the creation and enhancement of models through the creation of lists, modules, formulas, and documentation based on natural language prompts. All changes made by model builders are then tested before implementation to ensure fast but secure operations.
No. The purpose of Anaplan AI is to work with experienced model builders to assist them in their modelling, rather than to replace them. Anaplan AI helps users do the same actions much faster.
All functionalities of Anaplan AI inherit the existing workspace, role, and dimension-based security of the planning model. All data that AI agents are allowed to use is the data that the requesting user is allowed to view.
Organizations implement Anaplan AI solutions in finance, sales, logistics, workforce planning, merchandising, and operations. This is feasible because all business areas use the same planning solution and thus make coherent decisions.
Yes. Companies can add new capabilities such as CoModeler, Forecaster, or Agent Studio to existing environments without rebuilding planning models.
It depends on the company's priorities. For fast model creation, CoModeler is used; if a company is focused on AI governance, Agent Studio is a good starting point; when the aim is improving forecast accuracy, Forecaster should be implemented first.
Want to implement Anaplan AI in your organization? Talk to Polestar Analytics' Anaplan specialists about sequencing CoModeler, Agent Studio, and Forecaster inside your existing Connected Planning environment.