# AI assistants compared, six tools, six ways to bill Canonical URL: https://industrycontents.com/personal-ai-assistants-compared/ Language: en-GB Published: 2026-10-12 Structure updated: 2026-10-25 > Dated article snapshot. Canonical web page may receive later updates. Vendor figures retain their stated source-check dates. Illustrations are not product benchmarks. Published: 2026-10-12 · Structure updated: 2026-10-25. Vendor prices and source checks retain the dates stated below. ## Quick answer Lindy, Zapier Agents, Gumloop, Relevance AI, Writer and Manus use different billing units and sharing models. Choose by the work and controls required, then compare total cost per accepted task, including review and failed attempts. ## Key terms **AI assistant**: AI assistant is software that uses AI to help complete research, communication or connected-system tasks. **Billable unit**: Billable unit is the vendor-defined event or capacity deducted from a plan. **Cost per accepted task**: Cost per accepted task is the comparable total spend divided by task results that pass the acceptance criteria. ## AI assistants compared, six products and billing models | Product | Billing unit | Published price | Trial | | --- | --- | --- | --- | | Lindy | Active users plus pooled credits | Team from $29.99 per active user with 3,000 credits; $49.99 for 5,000. Top ups $10 per 1,000. Rechecked 25 October 2026 | $50 credit for 7 days | | Zapier Agents | Activities, separate from Zap tasks | Free for 400 activities a month. Pro costs $400 billed annually for 1,500 a month | Free plan | | Gumloop | Credits plus an orchestration fee | Pro from $37 a month with 20,000 credits and an 8% orchestration fee | 14 days | | Relevance AI | Actions plus Vendor Credits | Enterprise by quote on the pricing page. Documentation lists top ups of $80 per 1,000 Actions and $20 per 10,000 Vendor Credits | Free plan in documentation with 200 Actions and 1,000 Vendor Credits | | Writer | Seats plus credits plus agreement | Starter seat price not public. Enterprise is custom | 14 days, up to 5 users | | Manus | Credits | Pro from $20 a month for 4,000 credits, or $40 for 8,000 | Free plan with 300 daily credits | Caption: Compare six AI assistants and their starting prices. The table sets out Product, Billing unit, Published price, Trial. ## Check what each billable unit means Lindy pools credits across paid active users. Zapier Agents counts activities separately from Zap tasks, but its documentation lists agent sharing on Enterprise rather than Pro. Zapier is migrating Agents into AI by Zapier; use the applicable task rates for a migrated workflow. Gumloop meters credits plus an orchestration fee. Zapier sources checked 18 October 2026, [Agents usage rules](https://help.zapier.com/hc/en-us/articles/26559132765325-How-is-Zapier-Agents-usage-measured) and [migration guide](https://help.zapier.com/hc/en-us/articles/47402591569805-Migrating-from-Agents-to-AI-by-Zapier). The migration guide was rechecked 25 October 2026. According to its documentation, Relevance AI meters Actions and Vendor Credits separately, so a budget built on model costs alone will fall short. Writer's Starter plan allows up to five users during the 14 day trial. Manus credits are consumed by model tokens, virtual machines and third party APIs, and monthly credits do not roll over. ## AI assistant credits, unit price is not task cost An AI assistant’s price per credit is not its cost per task because credit consumption and accepted results differ across products. | Product | Plan | Credits | Price per 1,000 credits | | --- | --- | --- | --- | | Lindy | Team, $29.99 a month | 3,000 | about $10.00 | | Lindy | Team, $49.99 a month | 5,000 | about $10.00 | | Manus | Pro, $20 a month | 4,000 | $5.00 | | Manus | Pro, $40 a month | 8,000 | $5.00 | | Gumloop | Pro, $37 a month | 20,000 | about $1.85 before the 8% orchestration fee | | Relevance AI (documentation top ups) | Actions $80 per 1,000 and Vendor Credits $20 per 10,000 | Two meters | $80 per 1,000 Actions and $2 per 1,000 Vendor Credits | Caption: Compare credit prices without mistaking them for task costs. The table sets out Product, Plan, Credits, Price per 1,000 credits. These figures price each company's units. Which tool finishes a task most cheaply depends on how many units the task burns, and a credit buys a different amount of work at each company. Lindy says a job uses 2 to 2,500 credits depending on its size, and Manus says credits follow tokens, virtual machine time and third party APIs. Gumloop's credit definition sat behind a blank FAQ when we read it. For a meaningful comparison, run the same task in each product and record the credits consumed, the total cost and the result. If a Lindy deep work job uses 600 credits it costs about $6.00. A job that needs 6,000 Manus credits would cost $30 at $5 per 1,000, so the cheaper credit can still lose when the task burns ten times more of them. ## Watch for seat charges, overages and expiring credits Seat priced tools cost more as more people touch them, and credit priced tools cost more as tasks get longer. Only a trial shows how many tasks you will run. Zapier Pro is $400 billed annually for 1,500 activities a month, which its pricing page confirmed on 14 October. Relevance AI's public pricing page now lists only an Enterprise plan with sales led pricing, and third party figures for it disagree, so the tracker records none. ## Check Manus credit rules before choosing an alternative Manus lists Pro plans from $20 a month for 4,000 credits, or $40 for 8,000. Both work out to $5 per 1,000 credits at their starting allowances. That rate says little about the cost of a finished research task, because Manus says usage follows model tokens, virtual machine work and third party APIs. Monthly Manus credits do not roll over, and annual subscriptions still refresh their allowance monthly. Its documentation also says a running task can finish an action that pushes the balance below zero, while a new task cannot start at zero or below. Technical failures attributable to Manus may qualify for credit refunds. Confirm one detail before you buy add on credits. Manus's help centre and product documentation describe expiry and subscription requirements differently, so ask support whether purchased add ons stay usable after you cancel a paid plan. Our [Manus credit and subscription guide](/guides/manus-alternatives-cost/) carries the source links and worked examples. ## Match assistants to the task Separate one-off research, scheduled workflows and actions in connected systems. Test a research assistant on source quality and completeness. Test a recurring workflow on reliable scheduling, duplicate prevention and recovery. Test an assistant that changes records on permissions, approval and audit history. A single “best assistant” score hides those differences. ### Build an assistant task sample Choose two short tasks, two multi-step tasks and one scheduled task that resemble your working week. Define acceptance before the trial and record every retry. Do not convert a five-task sample straight into an annual forecast; repeat it with enough variation to expose expensive failures and unusual inputs. For invented monthly figures, a $100 subscription plus $40 extra usage and two review hours at $50 totals $240. With 80 accepted tasks it is $3 each. Another product at $150 with no extras but four review hours costs $350; with the same 80 accepted tasks it is about $4.38. The cheaper software can still be more expensive once the human work changes, so measure both. ## Choose the sharing model List who needs to create assistants, run them, approve actions or only read the output. Ask which roles require payment and whether usage belongs to a person or a workspace. Test what happens when a heavy user consumes a pool and when a staff member leaves with a connected account. Keep account access and workflow continuity separate. A shared login is not a substitute for supported team permissions. Ask who can revoke credentials, export configurations and see usage history, then price the tier that supports those duties. ### Leave room to move Save the task definition, source files, prompts, integrations and acceptance tests outside the tool. Before cancellation, test exporting results and preserving a scheduled workflow's state. The [context-handoff guide](/context-rot-switching-ai-tools/) covers the information another assistant needs to continue the work. ## Run these checks before choosing Assistant selection requires a defined billable unit, measured task consumption, paid-user rules and expiry terms. - What is one billable unit, whether a user, a step, a credit or an action? - How many of those does one of our real tasks use, measured on a trial? - Does anyone who touches the tool take a paid seat? - What happens at the limit, whether a pause, a top up or an automatic overage? - Do unused credits expire? For what happens when work moves between assistants, see [our guide to preserving context](https://industrycontents.com/context-rot-switching-ai-tools/). A lower subscription price can still leave a costly handoff. Lindy’s entry tier and active-user pricing were rechecked 25 October 2026. Other vendor figures retain their stated earlier check dates. ## Frequently asked questions ### Which AI assistant is cheapest for our team? The lowest starting subscription does not establish the lowest cost per completed task. Run the same workload, count paid users, check shared balances and include retries and review time. ### Can I compare one vendor's credits directly with another's? No. Each provider defines its billing unit differently. Compare observed cost per completed task and include subscription fees and extra usage. ### What should I check before enabling extra usage? Check spending controls, reset and expiry rules, paid-user eligibility and whether usage is shared. Confirm the applicable terms in your account before relying on a budget. [Compare Lindy, Zapier Agents and Manus team budgets](/ai-teammates-pricing/). Use the dated sources in this article for price context; the comparison method is an editorial framework, not a hands-on product benchmark. Depth update, 25 October 2026: the new task basket, sharing checks and review-cost calculations are editorial frameworks. Lindy and Zapier migration corrections in this series remain dated separately; this update does not claim new pricing verification for all six products. ## Sources - Pricing pages for [Lindy](https://www.lindy.ai/pricing), [Zapier](https://zapier.com/pricing), [Gumloop](https://www.gumloop.com/pricing), [Relevance AI](https://relevanceai.com/pricing) and [Writer](https://writer.com/pricing/), read 14 October 2026. - Manus Help Center, [membership pricing](https://help.manus.im/en/articles/11711111-what-is-the-current-membership-pricing-for-manus), read 14 October 2026. - Tracker entries for [Lindy](https://industrycontents.com/ic-lab/ai-agent-pricing-tracker/), Zapier, Gumloop, Relevance AI and Writer, each linking to the company's page. The Relevance AI documentation figures and the Manus credit rules come from earlier readings. See the AI Agent Pricing Tracker, [Lindy in detail](https://industrycontents.com/guides/lindy-pricing-seats-credits/) and Manus in detail. IC Lab · AI Agent Pricing Tracker Take the pricing rows into your budget. Compare source-linked rows in the tracker. Open the tracker and find the CSV export. The tracker offers a CSV export with a TalkHead signup. TalkHead is our weekly newsletter on AI in business. **Related guides** Lindy pricing · Manus alternatives · [Workflow automation pricing](/guides/workflow-automation-pricing/) · [Free AI agents](/guides/free-ai-agents/) · [Which AI agent, and what it costs](/guides/which-ai-agent-what-it-costs/) · AI Agent Pricing Tracker · [All AI agent guides](/guides/) ## Continue with the evidence **[The billing unit changes cost](https://industrycontents.com/ai-agents-breaking-saas-user-definition/)** An AI worker can consume credits without being a licensed employee. Our analysis examines how that changes enterprise software economics. **[Compare the pricing models side by side](https://industrycontents.com/ai-teammates-pricing/)** Compare billing units and linked company sources before budgeting.