Basic
$5 free credits
Then a 45% service fee on each top-up. Model costs at market rate.
The whole platform, self-serve. Ship your first agent today and only pay when you need more credits.
- Every feature in the platform, no gating
- LLM costs passed through at market rate
- Charged only when you top up. No monthly fee, no seats
- Basic support
No credit card required.
Pro
We build with youCustom
Lower service fee on credits, priced to your volume.
Everything in Basic, plus a team that onboards you and builds your agents with you, and the tooling to turn them into revenue.
- Everything in Basic
- Lower service fee on credits
- White-glove onboarding
- Forward-deployed engineer assigned to you, building your agents
- Advanced monetisation features
- Advanced analytics
- Priority support
Thirty minutes with the founding team.
How billing works
$5 free credits to start
Sign up and build. The first agents, traces and evals run on us.
Charged only at top-up
No monthly platform fee, no per-seat pricing. You pay when you add credits, and nothing in between.
LLM costs at market rate
Model usage is passed through at the provider's price. The service fee on each top-up is the only markup.
Everything included
Compare plans.
Basic is the full platform. Pro adds people, a lower fee and the tools to charge your customers for what you ship.
| Feature | Basic | Pro |
|---|---|---|
| Build | ||
| AgentsPrompt, model, memory, guardrails and tools configured in Studio. Nothing to host. | ||
| WorkflowsTools, code and agents on one canvas. Parallel and branching steps, every run timed and costed. | ||
| NetworksOne assistant out front, a handoff agent routing to specialists behind it. | ||
| Prompt variables & versioned releases | ||
| Dev → Staging → Prod environments | ||
| Visual builder in StudioProduct teams compose; engineers extend with custom tools when needed. | ||
| Models | ||
| 605 models from 60+ providersFilter by region, retention and training policy. Swap per agent in two clicks. | ||
| EU-hosted, zero-retention options | ||
| New models within 24 hours | ||
| Bring your own provider keys | ||
| Fallback models & automatic retries | ||
| LLM costsProvider pricing passed through at market rate. No markup on the model itself. | Market rate | Market rate |
| Tools & data | ||
| OpenAPI importPaste your spec and every endpoint becomes a tool. | ||
| MCP servers | ||
| 1,000+ integrationsSlack, Gmail, Sheets and more, with auth handled. | ||
| REST, GraphQL and SQL connectionsOAuth or API-key auth to your own internal APIs. | ||
| Tools scoped per agent | ||
| Memory & context | ||
| Memory (all 4 modes) | ||
| RAG & vector search | ||
| Tenant-scoped contextEach customer's agents see only that customer's data. | ||
| Where it runs | ||
| Chat SDKReact and web components to embed the assistant in your product. | ||
| REST APICall any agent or workflow from your backend. Streaming responses, typed schemas. | ||
| Webhooks | ||
| Scheduled tasks for your usersNightly reviews, weekly sweeps, hourly checks. Runs on a cron, results land in-product. | ||
| Slack, email, WhatsApp & SMS | ||
| CLI & MCP hostRun agents from the terminal or from Claude, Cursor and any MCP-compatible host. | ||
| Observability | ||
| Traces on every requestEvery model, memory and tool call timed and priced, with raw input and output. | ||
| Threads & filter by customer | ||
| Audit trail on every action | ||
| Evals | ||
| 14 scorers, code-based and LLM-basedRelevancy, faithfulness, hallucination and more. | ||
| Golden sets & experimentsProve a prompt or model change is better before you ship it. | ||
| Continuous scoring in production | ||
| Auto-optimiseFinds cheaper model setups that score just as well. | ||
| Costs & monetisation | ||
| Cost tracking per model, agent and customer | ||
| Live margin & value createdRevenue added, hours saved, churn caught, next to what each task cost. | ||
| Usage metering per customer | ||
| Advanced monetisationBilling rules per customer (markup, tiers, prepaid credits), Stripe and Paddle sync, entitlement API. | ||
| Advanced analyticsAdoption, retention and revenue dashboards across your customer base. | ||
| Security & guardrails | ||
| Guardrails per agent | ||
| Human-in-the-loop approvals | ||
| Per-tenant credential scopes | ||
| PII redaction across traces | ||
| EU or US hosting | ||
| Pricing & support | ||
| Free credits on signup | $5 | $5 |
| Service fee on creditsApplied to each top-up. You are only charged when you top up. | 45% | Lower, custom |
| Monthly platform fee | None | Custom |
| Support | Basic | Priority |
| Onboarding | Self-serve with docs | White-glove |
| Forward-deployed engineerAn engineer assigned to your account who builds your agents with you. | ||
| We build for youAgents, workflows and integrations built and shipped by our team. | ||
What am I actually charged for?
Credits. You start with $5 free. When you top up, a 45% service fee is applied to that top-up on Basic. That is the only charge: there is no monthly platform fee and no per-seat pricing.
How are model costs handled?
We pass LLM costs through at market rate, whatever the provider charges. Credits are spent at that rate as your agents run, and every trace shows the exact cost of each step.
What does the service fee cover?
Everything except the model: hosting, Studio, the chat SDK, workflows, memory, traces, evals, cost tracking, integrations and support. All of it is included on Basic.
Are any features gated on Basic?
The platform itself is not. Basic has every building block: agents, workflows, networks, all models, tools, memory, observability, evals and cost tracking. Pro adds services on top, plus advanced monetisation and analytics.
How is Pro priced?
Custom, based on your volume and how much of the build you want us to take on. Pro always comes with a lower service fee on credits than Basic. Book a call and we will put a number on it.