Pricing
We do not meter the behavior we want to encourage. Pick the plan that fits how your team runs agents today - memory operations stay unmetered for normal use.
For one developer, forever.
For individual developers exploring Spark on personal and project work.
Start freeIncludes
Billed annually - 5-seat minimum
For teams using coding agents daily and tired of re-solving the same problems.
Start trialEverything in Developer, plus
Platform and programme licensing.
For organizations running agents across teams, tools, models, and governed workflows.
Book a callEverything in Team, plus
No charge per memory call, retrieval, or agent run. If memory is working, your agents should use it constantly.
You budget around people, so do we. Large production-agent deployments are handled through Enterprise pricing rather than surprise usage fees.
What your agents discover becomes permissioned, auditable organizational memory that survives model and tool churn.
FAQ
No. Spark is priced per human contributor, not per memory call, retrieval, or agent run. Normal team usage is unmetered. Large production-agent deployments are handled through Enterprise pricing.
Because agents execute, but teams learn. Spark turns each contributor's decisions, corrections, and context into shared memory that future agents can reuse. Pricing by human contributor keeps that simple: your team's agents can work continuously without every run becoming a billable event.
Memory APIs sell storage and retrieval. Spark is the governed learning loop: what gets captured, permissioned, audited, promoted, and safely reused across your organization.
Yes, for self-serve teams. Enterprise workflows involving security, tenancy, policy, or programme memory start with a scoped pathfinder against written success criteria.
Tell us how your team runs agents and we will point you to the right plan - or scope a pathfinder if you are bringing Spark into an enterprise programme.
Book a call