Contents 14 sections
The best AI agent depends on the work: prompted research, business tasks, or workflow automation. AI agents plan and take actions across multistep tasks instead of only answering prompts.1
They can handle scheduling, research, data analysis, coding, and web-based work, though this ranking excludes coding specialists.1 OpenAI once documented 40 agent requests per month on ChatGPT Plus, but now says ChatGPT agent is no longer available2.
Comparison of AI agents: table and ranking criteria
ChatGPT agent, Lindy, Zapier Agents, n8n, Make, Relevance AI, Gumloop, and Microsoft Copilot Studio address different kinds of work. This ranking weighs task fit, documented usage, approval controls, and hosting. A credit allowance is a usage budget, not the same thing as a task or workflow execution.
The table lists each product in ranking order, with its type, free allowance, paid price, documented usage, and self-hosting status. The n8n Starter plan costs $20/month billed annually and includes 2,500 workflow executions per month3.
| Tool | Type | Free allowance | Paid plan per month | Documented usage allowance |
|---|---|---|---|---|
| 1ChatGPT agent | General agent | Discontinued; historically, paid plans only (no Free-plan access) | – | – |
| 2Lindy | No-code agent builder | $50 in free credits | Team plan: $29.99/month per active user | Team plan: 3,000 credits/month per active user |
| 3Zapier Agents | No-code agent builder | 100 tasks/month | Agents Pro: $33.33/month, billed annually ($400/year) | Agents Pro: 1,500 activities per month |
| 4n8n | Workflow automation | Free self-hosted Community Edition; no execution allowance specified | $20/month, Starter plan, billed annually (2,500 workflow executions/month) | 1,600 Assistant credits/month |
| 5Make | Workflow automation | 1,000 credits/month (Free plan) | $12/month billed annually (Core; 10,000 credits/month) | – |
| 6Relevance AI | No-code agent builder | Free plan: 200 actions/month + $2 in bonus vendor credits | Pro: $29/month billed monthly, or $19/month billed annually | Pro plan: 2,500 actions/month |
| 7Gumloop | No-code agent builder | Free plan: 5,000 credits/month | Pro: starts at $37/month | Pro includes 20,000 credits/month |
| 8Microsoft Copilot Studio | No-code agent builder | Free trial; credit allowance not specified | $200/month per tenant license | 25,000 Copilot Credits per pack/month |
Table data: 2,4,5,6,7,8,9,10,11,12,13,14
A dash (—) in the table means the vendor does not publish this value or the item does not apply to its offer.
A dash means the table assigns no comparable value for that criterion. The cited price is for Zapier Professional, which starts at $19.99 per month; it is not a Zapier Agents plan price. Zapier Copilot is listed among the features of the Free plan.6
ChatGPT agent is the task-based pick for people who want to assign prompted work. n8n suits teams that put self-hosted workflow automation first. Neither verdict claims the lowest cost or the widest approval controls.
ChatGPT agent
4.4/5- Type
- General agent
- Free allowance
- Discontinued; historically, paid plans only (no Free-plan access)
At a glance
- ChatGPT agent suits prompted research and document work; n8n suits teams designing automation workflows.
- OpenAI documented 40 monthly agent requests for ChatGPT Plus, although it now says ChatGPT agent is no longer available.
- ChatGPT Work lists more than 1,400 available plugins for retrieving material from other tools.
ChatGPT agent suits prompted research and document work; n8n suits teams designing automation workflows. ChatGPT Work brings tools, files, and work context together for drafting and collaboration.15
The choice is between assigning work in ChatGPT and building the process yourself. OpenAI documented 40 monthly agent requests for ChatGPT Plus, although it now says ChatGPT agent is no longer available2.
ChatGPT agent: features confirmed by the vendor
ChatGPT agent can draw on connected context to prepare documents, reports, analyses, presentations, and spreadsheets.15 ChatGPT Work lists more than 1,400 available plugins for retrieving material from other tools.15
One vendor example identifies employees whose training completion falls below 100%, then drafts Slack messages from a template.15 Those messages are drafts, not independently sent communications.
ChatGPT Work also describes chart updates and scheduled weekly report drafts.15 Its appeal is turning material scattered across tools and files into a usable draft. The examples do not show that every later action runs without review. If a weekly report relies on ChatGPT scheduled tasks, Plus allows five active tasks; Pro allows 1516.
ChatGPT agent: pricing, credits and limits
ChatGPT agent access varies by device and plan. The vendor lists desktop access across all plans, with browser and mobile access on Plus, Pro, Business, Enterprise, and Edu.15
That difference matters if you assign work from a phone or browser. The access description does not establish an agent-task or credit allowance for each plan.
ChatGPT agent: who is it best for?
ChatGPT agent is the best fit for turning connected work context into drafts and reports, rather than building a workflow. The vendor describes making a structured blog-post draft from product documents, research, and meeting notes.15 Teams that require self-hosting should look elsewhere. The same applies to buyers seeking a documented guarantee that drafted messages go out without approval.
Lindy
3.9/5- Paid plan per month
- Team plan: $29.99/month per active user
- Type
- No-code agent builder
- Free allowance
- $50 in free credits
At a glance
- Lindy’s Team plan starts at $29.99 per active user per month and includes 3,000 credits per user.
- Lindy reports a SOC 2 Type II audit and independent verification of GDPR and PIPEDA compliance.
- For personal data, confirm the processor relationship has the contract or other binding legal act required by GDPR Article 28.
Lindy builds business agents for client follow-up; ChatGPT agent focuses on prompted work using connected context. Compare the work involved, the pricing model, and how far the agent should act. Lindy’s examples cover inbox replies and payment reminders, but they do not establish typical customer results.4 Lindy’s Team plan starts at $29.99 per active user per month and includes 3,000 credits per user5.
Lindy: features confirmed by the vendor
Lindy describes agents that research a lead’s company, draft a reply in the user’s voice, and respond from an inbox.4 Its payment example follows overdue invoices through reminders, calls, and follow-up.4
The vendor says setup takes 2 minutes. Treat that as a vendor claim, not a measured setup time for every team.4
Lindy reports a SOC 2 Type II audit and independent verification of GDPR and PIPEDA compliance.4 Teams handling client information still need to assess those claims against their data and approval rules. For personal data, confirm the processor relationship has the contract or other binding legal act required by GDPR Article 2817.
Lindy: pricing, credits and limits
Lindy begins with the free credit allowance in the comparison table. Paid plans start at $49.99 per user per month, starting with 5,000 monthly credits per user; the team draws from a shared credit pool.4
Subscription and usage are separate costs. Adding users raises the subscription cost, while agent activity consumes credits.5
Lindy: who is it best for?
Lindy is the best fit for client-facing teams seeking inbox and payment-follow-up agents; those tasks anchor its vendor examples.4 It is a poor fit if the subscription must stay fixed as the team grows. The shared credit pool also calls for usage tracking. Buyers should decide whether that trade-off is worthwhile for their client work.
Zapier Agents
4.5/5- Paid plan per month
- Agents Pro: $33.33/month, billed annually ($400/year)
- Type
- No-code agent builder
- Free allowance
- 100 tasks/month
At a glance
- Zapier Agents lets teams build specialized agents connected to business data across 9,000+ apps.
- Zapier Agents Free includes 400 agent activities per month, and testing counts toward that allowance.
- Zapier lists automated-behavior allowances of 400 activities per month on Free and 1,500 activities per month on Pro.
Zapier Agents builds agents that act across connected apps; Lindy emphasizes client communication and follow-up in its examples. Compare their tasks, allowances, and administrative controls – not imagined unlimited autonomy. Zapier says agents can draw on company knowledge and act on command or through automation.18
Zapier Agents: features confirmed by the vendor
Zapier Agents lets teams build specialized agents connected to business data across 9,000+ apps.18 A vendor example has an agent read a website and prepare personalized sales emails.18
Zapier Copilot helps create agents, but its role differs from an agent’s task allowance.18 Zapier Agents Free includes 400 agent activities per month, and testing counts toward that allowance19.
Zapier Agents does not currently honor app and action restrictions set on a Zapier Enterprise account.6 That is a material administrative limit. Enterprise teams should not assume their existing account policy controls agent actions.
Zapier Agents: pricing, credits and limits
Zapier Agents usage is best assessed by agent activity, not the price of Zapier Copilot. Zapier lists automated-behavior allowances of 400 activities per month on Free and 1,500 activities per month on Pro. Those figures are not universal limits for every plan.6
Uploaded knowledge-source files, including TXT files, have a 100 MB limit.6,19
Zapier places new accounts in a 14-day trial of Zapier Professional without requiring a credit card.6 That is an account-plan trial, not a separately priced Zapier Agents subscription. Keep agent runs, Zap tasks, and Copilot charges distinct when estimating usage. Zapier Agents Pro includes 1,500 agent activities per month; those activities are separate from Zapier task usage19.
Zapier Agents: who is it best for?
Zapier Agents is the best fit for specialized agents working with business data across connected apps. Cross-app task execution is its documented focus.18 It is not the fit for Enterprise teams that must apply existing app and action restrictions to every agent.6
Choose it only when permitted agent actions match the team’s approval rules. Zapier Agents Pro includes 1,500 activities per month, so the planned work must fit that allowance19.
n8n
4.1/5- Paid plan per month
- $20/month, Starter plan, billed annually (2,500 workflow executions/month)
- Type
- Workflow automation
- Free allowance
- Free self-hosted Community Edition; no execution allowance specified
At a glance
- n8n lets teams define how an agent acts within a workflow.
- n8n can pause a workflow for a person’s decision instead of automating every action.
- n8n includes its preview Assistant in Cloud plans other than Enterprise.
n8n lets teams define how an agent acts within a workflow. Zapier Agents, by contrast, concentrates on agents working across connected apps. Hosting, workflow control, and usage accounting separate the products. n8n describes a platform for building agents and workflows users can follow.20 n8n Starter is hosted by n8n and includes 2,500 workflow executions per month9.
n8n: features confirmed by the vendor
n8n can pause a workflow for a person’s decision instead of automating every action. In one vendor example, it looks up an order after a refund request. A person then decides in Slack whether to issue the refund or open a ticket.20
That approval belongs to the example, not every n8n workflow. n8n documents nine approval channels for AI tool actions, including Slack, Gmail, Microsoft Teams, and built-in chat21.
n8n: pricing, credits and limits
n8n includes its preview Assistant in Cloud plans other than Enterprise. Starter’s allowance appears in the comparison table. Pro includes 4,000 or 9,600 Assistant credits per month, depending on plan size.9
Assistant credits measure use of the preview feature, not completed workflows. n8n Starter includes 1,600 Assistant credits per month, apart from its workflow-execution allowance9.
n8n lists saved-execution storage allowances of 2.5 GB, 25 GB, and 50 GB across its listed tiers.9 Storage matters if the team keeps execution records. Assess it separately from Assistant credits.
n8n: who is it best for?
n8n is the best fit for teams designing agent workflows with explicit human decisions. Its refund example shows how that handoff can work.20 It is not the fit for buyers who require the preview Assistant on Enterprise Cloud, which excludes it.9
Choose n8n when defining the workflow matters more than assigning an open-ended task.
Make
3.8/5- Paid plan per month
- $12/month billed annually (Core; 10,000 credits/month)
- Type
- Workflow automation
- Free allowance
- 1,000 credits/month (Free plan)
At a glance
- Make connects apps in workflows whose steps teams can control. n8n is another workflow-focused choice.
- Make connects to 3,500+ apps out of the box and supports custom integrations for services with an API.
- It lists encryption, single sign-on, role-based access, and audit logs alongside SOC 2 Type II and SOC 3 certifications.
Make connects apps in workflows whose steps teams can control. n8n is another workflow-focused choice. Compare integrations, usage accounting, and where approvals belong. Make counts usage when a module, or workflow step, performs an action.10
A standard module action uses one credit; some advanced features using Make’s AI Provider can use more22.
Make: features confirmed by the vendor
Make connects to 3,500+ apps out of the box and supports custom integrations for services with an API.23 It lists encryption, single sign-on, role-based access, and audit logs alongside SOC 2 Type II and SOC 3 certifications.23
Those controls help teams administer access. They do not mean every agent action requires human approval.
Make: pricing, credits and limits
Make includes 1,000 credits per month on its Free plan.10 Module actions generally consume credits; the listed error-handler modules do not.10
Make lists 10 MB of incomplete-execution storage, with a 2 GB maximum. Assess that storage apart from action credits.10,22
Make advertises a Start-up Plan offer for qualifying businesses with fewer than 20 employees. The offer provides 50% off Business.10 It is a conditional discount, not the standard paid-plan price.24
Make: who is it best for?
Make is the best fit for teams that can map agent work into connected-app workflow steps. Its integrations and module-based usage model suit that structure. It is less suitable if every module action must count as a flat-price task, or if approval must be assumed for every action. The choice turns on whether the team can define each step and budget the credits it consumes.
Relevance AI
4.4/5- Paid plan per month
- Pro: $29/month billed monthly, or $19/month billed annually
- Type
- No-code agent builder
- Free allowance
- Free plan: 200 actions/month + $2 in bonus vendor credits
At a glance
- Its workflow model separates assisted tasks from Level 3 autopilot, which routes incoming work to agents automatically.
- The image shows $84,210 in monthly AI usage, which should not stand in for a buyer’s subscription estimate.
- Another illustrated dashboard shows 129,480 checks run and 212 items sent to a person.
Relevance AI builds agents around business processes; Make centers on connected-app workflow steps. Compare the work assigned, its trigger, and the point where a person reviews an outcome. Relevance AI describes cloud-run agents triggered by business systems.25
Relevance AI: features confirmed by the vendor
Relevance AI describes an agent researching an account and contact before passing work to the next step.25 Its workflow model separates assisted tasks from Level 3 autopilot, which routes incoming work to agents automatically.25
Automatic routing does not mean every later action skips human review.
Relevance AI: pricing, credits and limits
Relevance AI uses a dashboard illustration to depict AI spending across model providers. It is neither a customer plan price nor an included allowance. The image shows $84,210 in monthly AI usage, which should not stand in for a buyer’s subscription estimate.25
Another illustrated dashboard shows 129,480 checks run and 212 items sent to a person. These are not plan limits or verified customer outcomes.25
Relevance AI: who is it best for?
Relevance AI is the best fit for teams designing cloud-run processes that route incoming business work to agents. That is the workflow its vendor describes.25 It is not a fit for buyers treating illustrated dashboard totals as predictable plan allowances. The distinction is business-process routing versus a fixed sequence of app actions.
Gumloop
3.4/5- Paid plan per month
- Pro: starts at $37/month
- Type
- No-code agent builder
- Free allowance
- Free plan: 5,000 credits/month
At a glance
- Gumloop shows an agent reviewing 38 accounts with 1 approval pending.
- The vendor also describes agents signing in through a saved login or 1Password connection without seeing the credential.
- Gumloop Pro is listed at $37 per month, without a stated billing commitment in the cited price.
Gumloop presents business-task agents that can pause for approval. Relevance AI puts more emphasis on routing incoming work to agents. Compare task type, approval points, and usage pricing. Gumloop’s account-review example ends with approval pending, rather than fully unattended work.26
Gumloop: features confirmed by the vendor
Gumloop shows an agent reviewing 38 accounts with 1 approval pending.26 The vendor also describes agents signing in through a saved login or 1Password connection without seeing the credential.26
Those examples support account monitoring and credential-mediated access. They do not establish approval as a requirement for every action.
Gumloop: pricing, credits and limits
Gumloop Pro is listed at $37 per month, without a stated billing commitment in the cited price.13 A pricing FAQ asks about exceeding 20,000 credits mid-month. That question alone establishes neither a plan allowance nor an overage rate.13
Separate the subscription price from credit-based usage when estimating total cost.
Gumloop: who is it best for?
Gumloop is the best fit for teams that want agents to review accounts and bring actions to a person for approval. Its vendor example shows that handoff.26 It is not a fit if the cited pricing information must establish a fixed monthly usage ceiling. Buyers need to define which agent actions require approval.
Microsoft Copilot Studio
3.9/5- Paid plan per month
- $200/month per tenant license
- Type
- No-code agent builder
- Free allowance
- Free trial; credit allowance not specified
At a glance
- Microsoft Copilot Studio connects agents to apps, sites, files, and business data through 1,500+ prebuilt data connectors and MCP servers.
- Microsoft describes default and custom model choices, including 11,000+ models through Azure AI Foundry.
- Microsoft advertises savings of up to 20% for an upfront purchase of Copilot Credit Commit Units.
Microsoft Copilot Studio builds agents connected to business systems. Gumloop instead illustrates agents bringing work forward for approval. Compare integrations, usage pricing, and the review required for each task. Microsoft describes creating agents through natural language or a graphical interface.27
Microsoft Copilot Studio: features confirmed by the vendor
Microsoft Copilot Studio connects agents to apps, sites, files, and business data through 1,500+ prebuilt data connectors and MCP servers.27 Microsoft describes default and custom model choices, including 11,000+ models through Azure AI Foundry.27
A wider set of connections and models does not imply human approval for every action.
Microsoft Copilot Studio: pricing, credits and limits
Microsoft Copilot Studio lists a $200 pack per month for 25,000 Copilot Credits.14 Pay-as-you-go billing charges for Copilot Credit capacity used at the end of the billing period, with no upfront license commitment.14
Microsoft advertises savings of up to 20% for an upfront purchase of Copilot Credit Commit Units. That conditional discount is not the standard pack price.27,28
Microsoft Copilot Studio: who is it best for?
Microsoft Copilot Studio is the best fit for teams building agents around connected business data while choosing committed credits or usage-based billing. It is not a fit for buyers who assume a monthly pack means unlimited activity. Integrations do not supply an approval step on their own. The team must forecast credit use and specify which actions need review.
Methodology and limits of this comparison
This AI-agent ranking weighs general agents, no-code builders, and workflow platforms by task fit, integrations, allowances, approvals, and hosting. It draws on public vendor pages, pricing pages, and documentation available at publication. It is not a hands-on performance test.
Ranking positions reflect documented fit for different work, not measured speed or reliability. Vendor examples describe possible workflows, not verified customer outcomes. Credits, tasks, and executions are different usage units. No product is the universal winner.
General AI agent, agent builder or workflow automation?
A general AI agent suits a task you describe; an agent builder suits a repeatable role you configure. Workflow automation suits a defined sequence of actions. Compare usage units after choosing the work: credits, successful tasks, and executions count different things. Zapier counts successful actions as tasks but excludes triggers; n8n counts a workflow run as one execution regardless of its steps.6,9
The table pairs each task with an agent type, example product, and documented usage unit.
| Task | Type that fits | Example tool | Usage unit |
|---|---|---|---|
| Draft a report from connected work context | General AI agent | ChatGPT agent | Plan-based access |
| Handle repeatable client follow-up | Agent builder | Lindy | 3,000 credits/month per active user on its starting paid plan |
| Act across connected business apps | Agent builder | Zapier Agents | 2 tasks per MCP tool call |
| Run a defined multistep process | Workflow automation | n8n | 2,500 executions on Starter |
| Build a visual app-connected workflow | Workflow automation | Make | Typically 1 credit per module action |
Table data: 5,6,9,10,15
Workflow design affects control and cost. Make lets builders set an agent’s goals, tools, and limits visually. Its advanced AI features can use more credits than a typical action.10,23
n8n lists 10,000 executions on Pro and 40,000 on Business. Its allowance scales by completed runs, not steps within a run.9
The practical choice follows the task. Use a general agent for prompted work or an agent builder for a recurring role. Choose workflow automation when each step and its cost need a definition.
Safety controls before an agent acts for you
AI agents need defined approvals, permissions, logs, data access, and spending limits before taking on business tasks.
- Human approvals should cover messages, record changes, and money commitments. Name the decision maker and pause before consequential actions.
- Agent permissions should cover only the apps and actions needed for the task. Microsoft Copilot Studio supports autonomous tasks and orchestration across agents, so set permissions by workflow rather than treating the platform as one boundary.27
- Execution logs should capture inputs, actions, outcomes, and reviewers. Relevance AI describes evaluations and monitoring alongside specialist agents and human participants; those records can help investigate exceptions.25
- Business-data access should stay within approved files and resources. ChatGPT Work draws context from team tools, files, and approved business resources, so choosing those sources is an access decision.15
- Spending limits should cover platform fees as well as usage. Gumloop Pro lists an 8% orchestration fee; account for it in the agent-activity budget.13
A safe agent workflow spells out what the agent can access and spend. It also states when a person must approve the next action.
What to check before you pay
AI-agent buyers should review free-tier expiry, usage caps, commercial rights, data use, and billing before paying. Lindy’s starting free credits expire after 7 days.5
The n8n Cloud trial includes 800 Assistant credits. Assess those apart from its workflow-execution allowance.9
- Free-tier access should support your intended device and task. ChatGPT Free and Go have limited desktop access to ChatGPT Work; Plus and Pro add browser and mobile access.29
- Usage caps should make clear what happens when credits run out. Make pauses scenarios until credits are added, while queued incoming webhooks remain subject to available storage.10
- Commercial use rights should cover how you intend to use agent-generated work. Review the applicable terms before publishing, selling, or delivering it to clients.
- Data-use terms should address connected files, customer information, and model training. Give agents access only to data your team may provide.
- Cancellation and billing terms should distinguish monthly payments from annual commitments. Make credits expire when the applicable subscription term ends.10
Total cost includes the subscription, team usage, and any extra credits – not just the entry price. Microsoft Copilot Studio offers prepaid packs, pre-purchase, and pay-as-you-go billing. Packs and pay-as-you-go have the same in-product features; the payment method differs.14
Compare each option against the same expected workload before committing.
Sources
- ↑10+ AI Agents You Can’t Miss in 2026 – They Feel Like the Future | God of Prompt https://google.com/goto?url=CAEScQHEIkXXy1wswRGZCZ3exFXXlCM2A1qE9-3Z5PbPrDJ3cbKhJFaTjCXDbiT2r9shobx61Nq8Onjb39XTW7uaxlgAhuyz7xUZRbAJT5LH2bwWwcaje3Hr-soVTT9dMdH86S1mfWNrM2ukZFzpXPBAFLOz&gsessionid=dEtYzYKKdzyY9mOdZwoTIt7a2ThaTK99kVOy1Cp1KJPS659O-zYLsw
- ↑OpenAI Help Center https://help.openai.com/en/articles/11752874-chatgpt-agent
- ↑n8n: n8n vs. Copilot Studio https://n8n.io/vs/copilot-studio/
- ↑Lindy: Hire an AI Employee That Gets Work Done https://www.lindy.ai/
- ↑Pricing – Lindy https://www.lindy.ai/pricing
- ↑Plans & Pricing | Zapier https://zapier.com/pricing
- ↑Host n8n | Deploy | n8n Docs https://docs.n8n.io/deploy/host-n8n
- ↑n8n Blog https://n8n.io/vs/azure-logic-apps/
- ↑n8n Plans and Pricing – n8n.io https://n8n.io/pricing/
- ↑Pricing & Subscription Packages | Make https://www.make.com/pricing
- ↑Pricing – Relevance AI https://relevanceai.com/pricing
- ↑Gumloop https://www.gumloop.com/blog/simplifying-gumloop-pricing
- ↑Pricing | Gumloop https://www.gumloop.com/pricing
- ↑Microsoft Copilot Studio Plans and Pricing https://www.microsoft.com/en-us/copilot/pricing/copilot-studio
- ↑ChatGPT Work: AI do projektów, zadań i pracy zespołowej https://chatgpt.com/work/
- ↑OpenAI Help Center: Scheduled tasks in ChatGPT https://help.openai.com/en/articles/10291617-scheduled-tasks-in-chatgpt
- ↑EUR-Lex: General Data Protection Regulation, Article 28 https://eur-lex.europa.eu/eli/reg/2016/679/ojv
- ↑Build AI teammates with Zapier Agents https://zapier.com/agents
- ↑Zapier, How is Zapier Agents usage measured?; Add your own data to an agent. ([help.zapier.com](help.zapier.com https://help.zapier.com/hc/en-us/articles/26559132765325-How-is-Zapier-Agents-usage-measured
- ↑AI Workflow Automation Platform – n8n https://n8n.io/
- ↑n8n Docs: Human-in-the-loop for AI tool calls https://docs.n8n.io/build/integrate-ai/ai-examples/human-in-the-loop-for-tools/
- ↑Make, Pricing & Subscription Packages. ([make.com](make.com https://www.make.com/en/pricing
- ↑AI Workflow Automation Software & Tools | Make https://www.make.com/
- ↑Make pricing. ([make.com](make.com https://www.make.com/pricing?pricing-country=US
- ↑Industrialize Your Revenue Supply Chain | Relevance AI https://relevanceai.com/
- ↑Gumloop: The enterprise AI agent platform https://www.gumloop.com/
- ↑Microsoft Copilot Studio | Create AI Agents https://www.microsoft.com/en-us/copilot/products/copilot-studio
- ↑Microsoft, Copilot Studio Licensing Guidance. ([microsoft.com](microsoft.com https://www.microsoft.com/licensing/guidance/Microsoft-Copilot-Studio
- ↑Plany i ceny | ChatGPT https://chatgpt.com/pricing/
Prices and terms checked at the source: November 2026.