Try Evoxiv on a single side project.
$0
no credit card · no time limit
- 1 agent · 1 product · 1 cronjob
- 1 seat
- 30-day Story history
- Community support · GitHub issues
for wiring up a single agent end-to-end
pricing
Flat pricing for a control plane.
You bring your own compute and your own LLM keys. We don't resell tokens or run agents on your behalf, so the price stays flat regardless of how hard you run your agents.
switch to annual to save ~17% · 2 months free
start for free
Solo developers who want every capacity limit removed.
$15 / month
or $150 / yr · 2 months free
- Unlimited agents, products & cronjobs
- 1 seat · 1 workspace — upgrade to Scale for teams
- Unlimited Story history
- Custom skills — markdown-defined
- Email support
when Evoxiv is doing real work for you
Teams that have moved past evaluation.
$20 / seat · month
2-seat minimum · $40 / mo entry
- Unlimited agents, products, cronjobs & workspaces
- 2+ seats · per-seat billing · no further caps
- Unlimited Story history
- Custom skills
- Priority email support
no Enterprise tier — contact us for invoiced billing
every plan includes
- Unlimited Story executions
- Built-in skills
- Per-agent Memory
- WebSocket updates
- GitHub App
- Slack integration
why flat
Three things we don't sell.
Evoxiv is a pure control plane. The expensive parts of an agent — compute, LLM tokens, sandboxes — never touch our infrastructure, so they never touch your invoice either.
— 01
Compute
The daemon that executes Stories runs on hardware you control — laptop, dedicated server, Cloud Run job. We never run it for you.
— 02
LLM tokens
You bring your own Anthropic, OpenAI, or other API keys. We never proxy LLM calls or add a markup on inference.
— 03
Sandboxes
The daemon's isolated workspace per Story executes on your hardware. Your code never flows through our infrastructure.
↳ paid
What you pay for
The control plane: API, web app, scheduling, history, integrations, real-time fan-out. Priced per seat and per workspace, not per Story.
compare
Every limit, on one page.
| free | pro | scale | |
|---|---|---|---|
| billing | |||
| Monthly price | $0 | $15 / mo | $20 / seat · mo |
| Annual price 2 months free | — | $150 / yr | $200 / seat · yr |
| Minimum commitment | none | none | 2 seats |
| capacity | |||
| Seats | 1 | 1 | 2+ · no cap |
| Workspaces | 1 | 1 | unlimited |
| Agents | 1 | unlimited | unlimited |
| Products | 1 | unlimited | unlimited |
| Cronjobs | 1 | unlimited | unlimited |
| Story executions | unlimited | unlimited | unlimited |
| Story history | 30 days | unlimited | unlimited |
| capabilities | |||
| Built-in skills memory, story, … | ✓ | ✓ | ✓ |
| Custom skills | — | ✓ | ✓ |
| Per-agent Memory | ✓ | ✓ | ✓ |
| WebSocket updates | ✓ | ✓ | ✓ |
| GitHub App · Slack | ✓ | ✓ | ✓ |
| support | |||
| Channel | community · GitHub issues | email · priority | |
counting rules
What counts toward each limit.
Limits are structural, not metered. Archiving frees a slot immediately — nothing carries over from one cycle to the next.
01 · agent
active rows only
Any non-archived row in agent. Archiving frees a slot immediately; archived agents are read-only.
02 · product
three live states
A row in product with status active, paused, or completed. archived doesn't count.
03 · cronjob
only when enabled
A row in cronjob with enabled = true. Disabled cronjobs are preserved and free to keep.
04 · seat
once invitation lands
Workspace members with role owner, admin, or member. Pending invites count once accepted.
billing operations
Upgrade, downgrade, cancel.
upgrade immediate
Free → Pro, Pro → Scale, or monthly → annual takes effect immediately. We prorate the difference against your current billing period.
downgrade at renewal
Scale → Pro, Pro → Free, or annual → monthly applies at the end of the current period. You keep the higher plan's capabilities until renewal.
over-limit grace 30 days
If a downgrade pushes you over the new plan's caps, existing resources keep running for 30 days. You can't create new ones until you archive enough to fit, then the workspace becomes read-only.
seat reduction next cycle
Removing a teammate on Scale reduces your invoice on the next cycle. Stripe handles the proration automatically.
annual switch prorated
Monthly → annual prorates against your current cycle. Annual → monthly takes effect at renewal. Annual is 10× monthly — effectively 2 months free.
cancellation self-serve
Cancel from the workspace billing page. The plan runs through the current period, then drops to Free. No fee, no exit interview, data preserved.
not included
And never will be.
Keeping Evoxiv as a pure control plane is what holds the price flat. The pieces below are by design outside our scope — they're also why your code never flows through our infrastructure.
- – compute
- You run
packages/daemonon a machine you control. Laptop, dedicated server, Cloud Run job — your call. - – llm tokens
- You bring Anthropic, OpenAI, or other API keys. We never proxy LLM calls; we don't even see the prompts.
- – sandboxes
- The daemon's isolated workspace per Story executes on your hardware. Filesystem and process isolation are yours to configure.
Can I run Evoxiv entirely self-hosted?
The daemon is already self-hosted by design — that's the whole point of the architecture. The control plane (API + web app) is a hosted service. Self-hosting the control plane is not offered.
What happens if I exceed the Story-history retention on Free?
Story rows older than 30 days are soft-deleted. Active Stories — status not done or cancelled — are never deleted regardless of age. Upgrading to Pro lifts the cap from that point forward.
Is there a discount for open-source or student use?
Not yet. If you have a public-good use case, open an issue and we'll look at it case by case. We don't publish a formal program because the cases tend to be unique.
Do you offer a free trial of Pro or Scale?
No — the Free tier is the trial. If Free isn't enough to evaluate, the $15 / month Pro plan can be cancelled within the first billing cycle for a full refund, no questions asked.
Do I pay for inference, compute, or storage on top of the plan?
No. The plan covers the control plane in full. Inference is billed by your LLM provider directly against your own API key. Compute (the daemon) runs on hardware you already pay for. Story metadata storage is included — there's no per-row charge.
How does the 2-seat minimum on Scale work?
Scale entry is $40 / month (2 seats × $20). Add seats freely — every additional teammate is +$20 / month, no further caps or tiers. Removing a teammate reduces your invoice on the next cycle, but you can't drop below 2 seats while on Scale; downgrade to Pro to go single-seat.
What if I outgrow Scale?
There's nothing to outgrow — Scale has no upper limit on seats, agents, products, or cronjobs. There is no Enterprise tier. If you need invoiced annual billing instead of card-on-file, contact us and we'll send a quote.
What payment methods do you accept?
All major cards via Stripe. For annual contracts at five or more seats we can issue an invoice payable by bank transfer — get in touch.
Wire up your first agent.
The Free tier is enough to take a real Story from idea to merged PR. Add a card the day you need a second product.