Your data never trains our models.
Thalium
Sign in Get started

PRICING

Start free. Scale as you grow.

Spark is free forever. Activate a 30-day trial on any paid tier — full institutional memory, compounding intelligence, no credit card required.

Spark

Free

Brain Instances

1

Invocations

500/mo

Memory

Session only

Role chain

5 roles

Support

Community

Evaluating Thalium or building a proof of concept.

Neuron

£29 /month

Brain Instances

3

Invocations

3,500/mo

Memory

Full institutional ring

Role chain

All roles

Support

Email

A focused internal tool or single-domain application in active use.

Lobe

£199 /month

Brain Instances

10

Invocations

30,000/mo

Memory

Full institutional ring

Role chain

All roles

Support

Priority email

A production SaaS feature or agency managing multiple client Brains.

Studio

£599 /month

Brain Instances

Unlimited

Invocations

100,000/mo

Memory

Full institutional ring

Role chain

All roles + custom

Support

Dedicated

A multi-product platform or high-volume agentic pipeline.

Enterprise

Custom

Brain Instances

Unlimited

Invocations

Negotiated

Memory

Sovereign deployment

Role chain

All roles + custom

Support

SLA-backed

Regulated industries, data residency requirements, or custom SLAs.

How the trial works

Your Brain learns before you pay.

01

Sandbox (unlimited)

Sign up free. Your Brain Instance processes every invocation, accumulates memory, and trains — silently. No output returned yet, no cost.

02

Activate your trial (30 days)

Click Activate in the dashboard. Your Brain starts returning real intelligence immediately. It already has context from the sandbox phase.

03

Subscribe or revert

At 30 days, subscribe to keep your Brain live. If not, it reverts to sandbox — memory preserved, nothing deleted.

Start free — activate your trial

Usage guidance

How many invocations do I need?

Neuron 3,500 invocations — ~115/day

A focused internal tool for one team. A single-domain support bot. A prototype in active development.

Lobe 30,000 invocations — ~1,000/day

A production SaaS feature serving hundreds of users. An agency managing 5–10 client Brain Instances.

Studio 100,000 invocations — ~3,300/day

A multi-product platform. A compliance tool processing hundreds of documents daily. High-volume agentic pipelines.

Need more? Overage invocations are billed at cost + a 30% coordination margin. Enable overage in your dashboard — you set the cap. Unused invocations do not roll over.

FAQ

Billing questions

What counts as an invocation?

One invocation is one call to the /v1/brain/{id}/invoke endpoint. The full role chain runs regardless of which roles are active for that intent type.

What happens when I hit my invocation limit?

On Spark, invocations are hard-capped. On paid tiers, you can enable overage billing — additional invocations at a per-invocation rate. You'll receive a notification at 80% of your allocation.

Can I change tiers mid-billing period?

Yes. Upgrades are effective immediately and prorated. Downgrades take effect at the end of the current billing period.

What is the difference between session memory and the institutional ring?

Session memory resets at the end of each session. The institutional ring is permanent — it persists and compounds across all invocations, indefinitely. Spark only has session memory.

Can I cancel at any time?

Yes. Cancel from your billing dashboard at any time. Your subscription remains active until the end of the paid period.

Is there a free trial of paid features?

Spark includes an optional 30-day live trial that activates full institutional memory. No credit card required to start.

What currency is billing in?

All prices are in GBP (£). Enterprise agreements can be denominated in USD or EUR on request.