
Taskade Free can support a small project evaluation, but do not budget for a recurring free AI allowance. The current official Free-plan guide says a new Free workspace starts with no AI credits: you earn credits or buy a pack. Before involving a client, check three separate constraints: workspace capacity, the balance available for AI work, and exactly what the client can access.
This pricing analysis uses official documentation and the live billing selector checked October 28, 2026. We did not purchase a plan, inspect a signed-in credit balance, send invitations or test access permissions. Our Taskade review covers turning a brief into reviewable tasks; this article addresses the narrower question of whether Free is sufficient to start a client project.
The current Free-plan guide, updated October 27, describes a permanent plan with no payment card required at signup. Projects and tasks are unlimited. That does not make every resource unlimited:
| Free-plan resource | Documented allowance | Buying implication |
|---|
| Workspace members | 2 | Count internal collaborators before inviting someone |
| Workspaces | 1 | Check whether separate client environments are essential |
| Genesis apps | 3 per workspace | Keep the number of apps distinct from projects |
| AI agents | 1, with 1 knowledge source | Define one narrow responsibility first |
| Automation runs | 10 per billing period | Forecast executions, not just workflow definitions |
| AI credits | None at signup; no Monthly Credits | Establish an actual balance before promising AI output |
| Files | 0.25 GB storage; 10 MB per file | Check the sample brief and attachments |
These allowances answer different questions. A consultant may fit within the member allowance and still lack credits for generation. Conversely, a remaining credit balance does not remove an app or member cap. Check the resource that the next action actually consumes.
Do not rely on an old signup-bonus figure. An older Help Center article still mentions up to 3,000 one-time credits. The newer guide and current FAQ state that Free begins at zero. Treat any account-specific promotional or earned balance as something to verify in your own account, not a guaranteed monthly entitlement.
Taskade's model and credit guide distinguishes Monthly Credits, which reset each billing cycle, from purchased or earned Extra Credits, which do not expire. It says Monthly Credits are used first. The model picker shows the current cost before an action; the guide points to Settings, then Plans & Billing, then Credits to inspect balances.
The same guide distinguishes building and AI work from ordinary use of an already published app. Visitors do not consume credits merely by running that published version. Do not extend that claim to every new agent run, AI automation step or builder edit: those involve AI work and need their own budget.
For a proposed evaluation, record the balance before one small request, inspect its result, then record the remaining balance. This is a suggested measurement, not a test we performed. Repeat only after identifying whether the next request is necessary. A long series of corrections can make a seemingly small assignment expensive in credits without producing an acceptable deliverable.
For example, begin with a fictional five-task brief and one acceptance criterion per task. Keep the source fixed while checking whether the output preserves scope. Do not predict how many client projects a credit pack will cover from a single request: prompt size, model choice and subsequent actions can change consumption.
Taskade's project-sharing guide documents sharing a single project by link or email without full workspace access. It distinguishes Editor, Checker and Viewer permissions. A Viewer can comment but cannot edit or complete tasks; Checker can complete tasks and requires a paid plan. Access can be revoked in the share settings.
For a client reviewing a draft, comments may be enough. Editing the whole task structure is a different requirement. Choose the narrowest access that supports the agreed review and inspect the sharing dialog before sending anything. A public link should not be treated as a private invitation simply because it is difficult to guess.
The roles and permissions guide separates workspace roles, per-project sharing and App Users on a published Genesis app. An app visitor is a different audience from a teammate who builds inside the workspace. Decide which kind of collaboration the client actually needs before interpreting the two-member allowance.
Use a harmless sample to check access with the intended recipient role and a separate session. Confirm that the reviewer sees the intended project, cannot reach unrelated work, and can perform only the agreed actions. Remove the sample access afterward. These are proposed acceptance checks, not verified results from a live Taskade account.
The live pricing page showed the following USD prices when we switched between monthly and yearly billing. These are plan prices with included people, not a per-person multiplication. Annual amounts require yearly billing; confirm taxes and renewal terms at checkout.
| Plan | Month-to-month | Annual-billing equivalent | Included workspace people |
|---|---|---|---|
| Free | $0 | $0 | 2 |
| Pro | $20/month | $10/month; $120 billed yearly | 10 |
| Business | $50/month | $25/month; $300 billed yearly | Unlimited |
The same pricing matrix lists Pro with 50,000 monthly AI credits and 1,000 monthly automation runs; Business has 150,000 and 10,000 respectively. Those are workspace resources. The current Free guide explicitly says the credit and run allowances are shared rather than multiplied by the number of members.
Compare the resource that blocks the job with the cost of resolving it. If all you need is occasional additional AI work, inspect the credit-pack option and its actual price before committing to an annual subscription. If the client needs a collaboration feature Free lacks, more credits alone will not provide it. We did not buy a pack or verify its checkout price.
Stay on Free for a bounded evaluation when its collaboration and capacity limits fit and you have established a usable credit balance. A manually maintained project is also a valid baseline; AI generation should earn its place through reviewable output.
Evaluate Pro when a recurring workload, additional collaborators or higher allowances create a clear requirement. Evaluate Business when its team controls or larger resource pool address a documented need. Avoid choosing a tier solely because a demonstration looks impressive: the practical result is an accepted project and controlled access, not the number of features enabled.
Open the Taskade listing, verify current provider terms, and write down three things before proceeding: who needs access, which allowance will constrain the work, and what evidence will make the evaluation successful. If the output still needs extensive correction, paying for a larger quota is not yet evidence that adoption makes sense.