Free
$0/forever
Experience the full Donkit magic.
3,000 credits · 1 seat
- Full agent builder
- Agents, workflows & scheduling
- API access & integrations
- Long-running tasks
- Limited execution history
- Community support
Free plan to get started, Pro and Team plans for scale. A credit is one unit of agent work and covers everything a run needs.
Experience the full Donkit magic.
3,000 credits · 1 seat
Run agents on real workloads.
20,000 credits · 1 seat
Operate a fleet of agents.
80,000 credits · 1 seat
Build agent systems with your team.
120,000 credits shared · 3 seats · +$39 & 30,000 / extra seat
Deploy AI workforce at company scale.
Custom committed usage · Custom seats
| Free | Pro | Pro Plus | Team | Enterprise | |
|---|---|---|---|---|---|
| Capacity | |||||
| Monthly credits | 3,000 | 20,000 | 80,000 | 120,000 shared | Custom |
| Included seats | 1 | 1 | 1 | 3 (+ $39 / seat) | Custom |
| Top-ups | — | Yes | Yes | Yes | Custom |
| Build | |||||
| Core agent builder | |||||
| Agents & workflows | |||||
| Scheduling | |||||
| Frontier models | — | ||||
| API access | |||||
| Integrations | |||||
| Long-running tasks | |||||
| Collections, memory, documents | |||||
| Observe | |||||
| Execution history | Limited | Standard | Extended | Extended | Custom |
| Tracing | Basic | ||||
| Observability | — | Basic | Advanced | Advanced | Custom |
| Auto build QA | |||||
| Evals | — | ||||
| Agent optimization | — | — | |||
| Collaborate | |||||
| Shared workspace | — | — | — | ||
| Shared billing | — | — | — | ||
| Roles & permissions | — | — | — | ||
| BYOK | — | — | — | ||
| Govern | |||||
| SSO · SAML · SCIM | — | — | — | — | |
| Audit logs | — | — | — | Standard | Advanced |
| Custom DPA | — | — | — | — | |
| Private / on-prem | — | — | — | — | |
| SLA & dedicated support | Community | Priority | Priority | Dedicated | |
Tokens are how model providers bill the infrastructure underneath — they say nothing about the work an agent actually did for you. A credit is one unit of agent work, and it bundles everything the run needed: model reasoning, tool calls, browser sessions, data access, memory, retries, scheduling and runtime. One predictable number on the invoice, no per-provider math, and the pricing stays sane when you swap models.
Everything that makes Donkit Donkit. The full agent builder, agents and workflows, scheduling, API access, integrations, and long-running tasks — all of it. The only thing Free constrains is monthly work capacity (3,000 credits) and how far back your execution history goes. No credit card required, and you can stay on Free as long as you like.
Nothing breaks and you don't get a surprise bill. Scheduled and ad-hoc runs that would push you past your limit pause cleanly with a clear status, your existing data and agents stay intact, and you choose what to do next: top up for a one-off boost, upgrade to a larger monthly plan, or wait for the period to reset. We never auto-charge overages.
Monthly plan credits reset at the end of each billing period — they're sized for steady usage, not stockpiling. Purchased credits behave differently: top-ups never expire, so you can buy a bigger pack during a busy month and spend it down whenever you need it.
Yes. Team is fully self-serve. It starts at 3 seats with 120,000 shared credits, $117 / month. You can add seats any time from billing — each extra seat is $39 / month and adds 30,000 credits to the shared pool. No sales call, no minimum commitment, cancel any time.
Available on Team and Enterprise. Connect your own provider key (OpenAI, Anthropic, Azure, and more) or a private OpenAI-compatible endpoint on the Your models page, and your model usage is billed directly by that provider at their rates — Donkit stops charging credits for those tokens. Credits still cover everything else around the model.
Yes, on Enterprise. Donkit can run in your own cloud account, in a dedicated single-tenant environment, or fully on-prem. Comes with a custom DPA, SSO/SAML/SCIM, advanced audit logs, and a dedicated TAM/CSM. Reach out to sales — [email protected].
No. Soft fair-use limits exist to keep the platform healthy, but we don't use API rate limits or task duration as pricing levers. Long-running tasks are first-class on every plan — including Free.
Yes. Upgrades take effect immediately and the new credit allowance is prorated to the rest of the period. Downgrades and cancellations take effect at the end of the current billing cycle — your agents keep running on the old plan until then, and you can reactivate before the period ends without losing anything.
All major credit and debit cards via Stripe. Invoicing in USD and EUR is available on Team and Enterprise; Enterprise also supports wire / ACH and custom net terms. Tax is handled automatically based on your billing address.