A8gent
GlossaryAutomation mechanics

What is copilot Credit?

Also called: copilot studio credits, copilot studio messages, copilot studio message, microsoft copilot credits

A Copilot Credit is the common billing currency across Microsoft Copilot Studio, which replaced the old message unit on 1 September 2025, and it is consumed at a published rate per response type: 1 for a classic answer, 2 for a generative answer, 5 for an agent action and 10 for tenant graph grounding.

Written by A8gent
Contents
  1. At a glance
  2. What a Copilot Credit is
  3. The published rate card
  4. The arithmetic Microsoft itself publishes
  5. Capacity, expiry and the 125% cutoff
  6. What you should know
  7. FAQ
  8. Related terms

At a glance

Unit
Common billing currency across Copilot Studio capabilities
Replaced
Messages, on 1 September 2025, at the same pack quantity and rate
Classic answer
1 credit
Generative answer
2 credits
Agent action
5 credits
Tenant graph grounding
10 credits
Agent flow actions
13 credits per 100 actions
GenAI voice
35 credits per minute; premium GenAI voice 75 per minute
Rollover
None - capacity enforced monthly, unused credits do not carry over
Cutoff
Custom agents disabled at 125% of prepaid capacity

What a Copilot Credit is

A Copilot Credit is the common currency across Microsoft Copilot Studio capabilities. Microsoft's licensing documentation is explicit about the history: starting on 1 September 2025, the common currency for agents changed from messages to Copilot Credits, with no change in the quantity per prepaid pack or the pay-as-you-go rate. If you are reading older material that talks about Copilot Studio messages, it is describing the same meter under its previous name.

Microsoft describes credits as measuring the time and effort an agent needs to retrieve information, respond to prompts and use actions or custom skills, with the number consumed depending on the complexity of the task. In practice that resolves to a fixed published rate card per feature type.

The published rate card

As documented in August 2026, the billing rates are: classic answer 1 credit; generative answer 2 credits; agent action 5 credits; tenant graph grounding for messages 10 credits; agent flow actions 13 credits per 100 actions. AI tools are rated by tier: basic text and generative AI tools 1 credit per 10 responses, standard 15 credits per 10 responses, premium 100 credits per 10 responses. Content processing tools are 8 credits per page.

Voice has its own table: classic voice on classic orchestration is 10 credits per minute, GenAI voice is 35 credits per minute, and premium GenAI voice is 75 credits per minute, each including core agent activity.

Reasoning models are billed twice over. Microsoft documents the total as the feature rate for the operation plus the text and generative AI tools (premium) rate for the reasoning model's token usage, at 10 credits per 1,000 tokens. A generative answer that uses a reasoning model is therefore 2 credits plus the premium token charge, not 2 credits.

The arithmetic Microsoft itself publishes

Microsoft's own worked examples are the clearest guide. A support agent averaging four classic answers and two generative answers per conversation, at 900 conversations a day, is [(4x1)+(2x2)] x 900 = 7,200 credits per day. An order-processing agent making four action calls is 4x5 = 20 credits per run. A tenant-graph-grounded agent at four generative answers and four grounded responses for 100 unlicensed users is [(4x2)+(4x10)] x 100 = 4,800 credits per day.

That third example is the one to internalise. Tenant graph grounding at 10 credits is five times a generative answer, so turning it on multiplies the cost of every grounded response. Microsoft notes a single complex prompt can cost 12 credits when grounding and generative answers combine.

Capacity, expiry and the 125% cutoff

Copilot Studio enforces purchased capacity monthly and unused credits do not carry over. Capacity is pooled across the tenant and can be allocated per environment. Enforcement triggers at 125% of prepaid capacity, at which point custom agents are disabled - ongoing conversations finish, but new invocations are rejected until capacity is increased or resets. Users see either "There is a billing issue" or "This agent is currently unavailable. It has reached its usage limit."

Agent flows enforce differently and earlier: when prepaid capacity is fully consumed, new agent flow runs are blocked while the parent agent keeps working for non-flow interactions.

What you should know

The largest single discount available is a Microsoft 365 Copilot licence. Microsoft documents that for a licensed user, employee-facing agent usage - classic answers, generative answers, tenant graph grounding, agent actions - is zero-rated. The 4,800-credit example above counts only the 100 unlicensed users; the 50 licensed ones cost nothing. Whether your audience is licensed is therefore a bigger cost variable than agent design.

Note also that Power Automate cloud flows are not billed in Copilot Credits - they use Power Automate licensing. Only agent flows managed inside Copilot Studio draw on this meter.

Copilot Credit FAQ

What is a Copilot Credit?

The common billing currency across Microsoft Copilot Studio. Microsoft describes it as measuring the time and effort an agent needs to retrieve information, respond and use actions, and publishes a fixed rate per feature type rather than leaving it to interpretation.

Are Copilot Studio messages the same as Copilot Credits?

They are the same meter under a different name. Microsoft documents that on 1 September 2025 the common currency for agents changed from messages to Copilot Credits, with no change in the quantity per prepaid pack or the pay-as-you-go rate.

How many credits does a generative answer cost?

Two, as published. A classic answer is 1, an agent action is 5, and tenant graph grounding is 10. A single complex prompt that uses grounding plus a generative answer is 12, which is the example Microsoft gives in its own documentation.

What happens if I run out of Copilot Credits?

Microsoft allows a grace band, then enforces at 125% of prepaid capacity by disabling custom agents. Ongoing conversations finish; new invocations are rejected with a billing or usage-limit message until capacity is increased, reallocated, or the month resets. Pay-as-you-go environments are not cut off because overage bills to Azure.

Do Microsoft 365 Copilot licensed users consume credits?

For employee-facing scenarios, no. Microsoft documents classic answers, generative answers, tenant graph grounding and agent actions as no charge when the agent operates under an authenticated Microsoft 365 Copilot licensed user's identity, subject to fair usage limits. This is the single largest variable in a Copilot Studio bill.

Do reasoning models cost more Copilot Credits?

Yes, on two meters. Microsoft bills the feature rate for the operation plus the text and generative AI tools (premium) rate for the reasoning model's token usage, at 10 credits per 1,000 tokens. A generative answer using a reasoning model is therefore well above its headline 2 credits.

Do Power Automate flows use Copilot Credits?

Power Automate cloud flows do not - they are licensed under Power Automate and are not subject to Copilot Studio billing or enforcement. Agent flows managed inside Copilot Studio do, at 13 credits per 100 actions.

Primary source

Microsoft Learn - Copilot Credits billing rates

Checked August 2026. Anything priced or rate-limited moves, so verify against the source before you budget.

Want this built rather than explained?

Knowing what copilot credit means is one thing; having it running in your business is another. We scope one workflow at a time, quote a fixed price in writing after a free consult, build it, and hand it over with documentation you own.