Quick Answer
Microsoft published its Copilot Credits Guide for September 2026, formalizing usage-based billing across Copilot Cowork, Studio, Dynamics 365, and Power Platform. Work IQ APIs — the runtime intelligence layer that grounds agents in enterprise context — cut token use by up to 80% for agentic solutions. Credits are pooled at the tenant level and purchased through the Copilot Studio pay-as-you-go meter or the Copilot Credit Pre-Purchase Plan (P3).
Last verified: Sep 16, 2026.
At a glance
- Guide published: September 2026
- Model: Usage-based Copilot Credits, pooled at tenant
- Work IQ token reduction: Up to 80%
- Purchase options: Pay-as-you-go meter, Copilot Credit Pre-Purchase Plan (P3), CSP partners, MACC
- Workloads covered: Copilot Cowork, Copilot Studio, Dynamics 365, Power Platform, Work IQ APIs
- GPT-6 Astra availability: Foundry GA since September 3, 2026
- OpenAI subprocessor: Available since July 9, 2026
What the Copilot Credits model is
Copilot Credits are the common currency for Microsoft's usage-based AI billing model. The number of credits consumed per response, action, or operation depends on the complexity of the task performed — multi-step agent flows consume more credits than single-shot chats. Credits are pooled at the tenant level, so an organization's total cost is the sum of credits consumed across all supported experiences (Microsoft Copilot Credits Guide, September 2026).
This is a meaningful shift from the prior per-seat model. With Copilot Credits, Microsoft aligns with the broader industry trend toward consumption-based AI billing — the same direction Salesforce is taking with Agentforce outcome-based pricing (Salesforce investor webinar, September 1, 2026). For enterprises with variable agent workloads, usage-based billing can be substantially cheaper than per-seat. For enterprises with heavy always-on workloads, per-seat can still win.
What is Work IQ
Work IQ is the workplace intelligence layer that grounds Copilot in enterprise context. Built on chat, context, tools, and workspaces, Work IQ continuously transforms signals across Microsoft 365 and business systems into agent-ready intelligence for grounding, reasoning, and permission awareness. The result is a platform that "enables agentic intelligence, speed, and efficiency at enterprise scale" with security and governance built in (Microsoft Copilot Credits Guide, September 2026).
The most consequential architectural decision: by shifting more AI processing into the Work IQ runtime, tokens needed for context retrieval and tool use are reduced by up to 80%, thereby reducing Copilot Credit consumption in agentic solutions. For a high-volume Copilot Studio deployment, that 80% reduction on the context-retrieval portion of agent execution translates into materially lower credit burn.
What Copilot Cowork does
Copilot Cowork is Microsoft's flagship agentic system for autonomous multi-step work. Describe the outcome you want, and Cowork does the work — sending emails, scheduling meetings, creating documents, posting in Teams, and managing files, while the user stays in control. Cowork is grounded in Work IQ and operates within Microsoft 365's security and governance boundaries, executing multi-step work autonomously at enterprise scale (Microsoft Copilot Credits Guide, September 2026).
From a credits standpoint, Cowork is the workload that drives the highest consumption because multi-step agent flows burn more credits than chat completions. Customers evaluating Cowork should plan for credit consumption patterns similar to a Salesforce Agentforce production deployment, not a ChatGPT-style chat subscription.
How Work IQ APIs bill
Work IQ APIs are billed separately from the per-user Copilot subscription. Microsoft Copilot is natively built on Work IQ and does not need to leverage Work IQ APIs — there are no incremental charges for Work IQ accessed in Microsoft Copilot experiences (chats, AI in Word/Excel/PowerPoint/OneNote/Outlook/Teams, agents natively included in Microsoft Copilot such as Researcher, Facilitator, and Analyst). The usage charges apply only to Work IQ API access in custom agents and AI solutions (Microsoft Copilot Credits Guide, September 2026).
For organizations building custom Copilot Studio agents or integrating Work IQ into Power Platform workflows, Work IQ API access is a separate line item from the per-user Copilot license. Budget accordingly — Work IQ API usage is what powers the "grounded in your data" claim that custom agents make, and the credit cost scales with the depth of grounding.
GPT-6 Astra in Microsoft Foundry
GPT-6 Astra is generally available in Microsoft Foundry since September 3, 2026. Astra is offered with Standard and Provisioned Throughput deployment options in both Global and US Data Zone geographies. Standard provides pay-as-you-go flexibility for variable demand, while Provisioned Throughput provides dedicated model-processing capacity for workloads requiring consistent latency and guaranteed throughput (Microsoft Azure blog, September 3, 2026).
Astra availability through Foundry is part of a broader pattern: Microsoft has been expanding model choice in Foundry since adding OpenAI as a subprocessor in July 2026. Foundry customers can select from OpenAI models operated by Microsoft (Azure OpenAI), OpenAI-operated models through OpenAI as subprocessor, and direct Azure models from Cohere, DeepSeek, Meta, Mistral AI, and xAI. The September 2026 Copilot Credits Guide formalizes the billing path for all of these (Microsoft Foundry, September 4, 2026).
What enterprise buyers should do
Three actions for organizations deploying Microsoft Copilot at scale.
- Map current Copilot workloads to credit consumption patterns. Chats consume few credits; Cowork, Studio agents, and Dynamics 365 AI consume many. The pricing model rewards shifting stable, predictable workloads to the per-user subscription and isolating bursty agent workloads to credit consumption.
- Provision the Copilot Credit Pre-Purchase Plan (P3) via MACC. This gives committed-use pricing for credit consumption and is the recommended procurement path for enterprises with stable Azure commitments.
- Plan Work IQ API access as a separate budget line. Custom agents and AI solutions using Work IQ APIs are billed separately from the per-user subscription. The 80% token-reduction figure applies to agent execution but does not change the billing model.
What to watch next
Three near-term datapoints. First, the October 2026 Microsoft Ignite conference will surface the next set of Copilot Credits pricing details and workload expansions. Second, the rollout of Copilot Cowork into general availability beyond enterprise tenants will change the credit-consumption profile materially. Third, any expansion of the per-user Copilot subscription to include more Copilot Credits by default would shift the breakeven point between per-seat and usage-based pricing for many enterprises.









