PatchBuddy. Get started
Pricing

Three tiers. Pay for what you ship.

A monthly subscription that includes a credit balance, seats, and an active-project allowance. AI usage on top, charged per token at the rate shown in your dashboard. Studio earns 10% off tokens, Agency earns 20%.

Annual billing saves 15% across the board. No hidden vendor markups. No procurement detour to find out what something costs.

Switching to annual locks the rate for 12 months and saves up to £900 a year.

All prices in £, ex VAT.

Solo

For freelance integration devs

£60 / month

Billed monthly

Save £108/yr on annual →

What's included
  • £30 credit per month
  • 1 seat
  • 3 active projects
  • Standard token rates
Choose Solo
Most agencies pick this
Studio

For small integration teams

£250 / month

Billed monthly

Save £450/yr on annual →

Everything in Solo, plus
  • £50 credit per month
  • 5 seats included
  • Additional seats £50/month each
  • 10 active projects
  • 10% discount on token usage
Choose Studio
Agency

For full integration agencies

£500 / month

Billed monthly

Save £900/yr on annual →

Everything in Studio, plus
  • £100 credit per month
  • Unlimited seats — no per-seat fee
  • Unlimited active projects
  • 20% discount on token usage
Choose Agency
Every tier ships with
Multi-tenant org access Audit log Encrypted Patchworks credentials Multi-model routing Knowledge fork All future feature shipments
Top-ups

Burnt through your credit?

Top up any amount, any time. Top-up credits never expire and stack on top of your monthly subscription credit. Studio and Agency tier discounts apply to every token spent — including top-up usage.

£25
credits
Most common
£100
credits
£500
credits
Or any custom amount from £5 minimum .
AI usage

Per-token. Visible. Live.

Every chat shows the cost as you go: input tokens, cache hits, output tokens, model selection, time-per-iteration. The price you see in-app is the price you pay. No surprise overruns at month end.

Cache hit ratios on most agency chats run high. Repetitive build work re-uses the same context, and our prompt-cache integration keeps costs in single-digit pence per turn for the typical iteration.

Live turn cost
Input
£0.0021
Cache hit
94%
Output
£0.0019
Tools
£0.0002
Total £0.0042
Model pricing Live from API

Per-model rates. All in £.

Below are the standard rates — what every Solo customer is charged per token. Studio applies 10% off these numbers, Agency applies 20% off. The dashboard shows the rate that actually ran on each turn so there are never any surprises.

Provider Model Context tokens Suggested use Input £ / 1M tokens Output £ / 1M tokens
Anthropic
Claude Haiku 4.5 200K Lightning fast + cheap — high-volume / short turns. £1.27 £6.33
Claude Opus 4.8 1M Deepest reasoning, slower — for complex planning + tricky decisions. £6.33 £31.64
Claude Sonnet 4.6 1M Balanced reasoning + speed — solid default for most chat. £3.80 £18.98
OpenAI
GPT-5 mini 272K Fast + affordable OpenAI — general-purpose everyday tasks. £0.32 £2.53
OpenAI GPT 5.4 1.1M OpenAI flagship — versatile across writing, code, analysis. £3.16 £18.98
DeepSeek
Deepseek Reasoner 131K Strong reasoning at low cost — budget-friendly thinking model. £0.35 £0.53
Moonshot
Kimi K2.6 (Moonshot) 262K Big-context Chinese model — long transcripts, multilingual chat. £1.06 £4.47
Mistral
EU-hosted
Large 3 262K GDPR Compliance + Slower but handles more complex tasks. £0.63 £1.90
Medium 3.5 32K GDPR Compliance + Balanced reasoning + speed £3.42 £10.25
Small 4 32K GDPR Compliance + fast + cheap — high-volume / short turns. £0.13 £0.38
Google
Coming soon · not currently routed
Cache savings

Where the provider supports prompt caching, cache hits charge at roughly 10% of the input rate. Most agency workloads run 80–95% cache hit on iterative builds. Your dashboard shows the cache portion of every turn.

No FX surprises

Rates are quoted in £ and charged in £. We absorb the USD→GBP conversion volatility so your monthly forecast doesn't move with the exchange rate.

Live data

This table is pulled live from our pricing API at build time. When we add a model or rates shift, the table refreshes on the next deploy. Your in-app dashboard always shows the rate active for every turn you've run.

Prices include all markup — they are the customer-facing rates per million tokens.

FAQ

Pricing, in detail.

What counts as an "active" project?

Anything not yet archived. When a build's done, archive it from the org dashboard — it keeps its data, chats, screenshots, and metrics indefinitely, it just doesn't count against your tier's active-project limit. The cap is a workflow nudge, not a wall.

What happens when I exceed my monthly credit?

AI calls stop until you top up. Subscription accounts can't run a negative balance — when subscription credit + any top-up balance hits zero, calls pause. Topping up is instant from the dashboard (£5 minimum) and the next message goes through immediately. Any unused top-up credit carries over forever.

Does my monthly subscription credit roll over?

Yes. Unused subscription credit carries forward into the next month. Quiet month? You build a buffer. Busy month? You burn it down. Top-up credit you've separately purchased never expires either.

Can I add seats to Studio mid-month?

Yes. Each additional Studio seat is £50 / month, pro-rated to the day you add them. Drop seats whenever a contractor finishes — billing adjusts on the next cycle.

Is there really no per-seat fee on Agency?

Right. Agency is a flat £500/month (£425 on annual) for unlimited seats and unlimited active projects. The token discount and the project ceiling are what you're paying for; seats just come along.

How does the token discount work?

Studio knocks 10% off every per-token rate in the table further up this page; Agency knocks 20% off. The discount applies to subscription-credit usage AND any top-ups you buy. The dashboard shows the post-discount rate on every turn so there's no end-of-month surprise.

What's the annual saving in real money?

15% across the board. Solo £612/year (£108 saved). Studio £2,550/year (£450 saved). Agency £5,100/year (£900 saved). Paid up-front, locks the rate for the year.

Can I downgrade tiers?

Yes, at the end of a billing period. If you're going Studio → Solo and you currently have more than 3 active projects or more than 1 seat, archive projects / remove seats first to fit Solo's caps. We'll prompt you in-app.

What about really big agencies — 30+ developers, dozens of clients?

Agency tier covers it — seats and projects are unlimited, and the more you use, the more the 20% token discount returns to you. If your usage profile is genuinely outside the curve, get in touch and we'll look at a custom volume arrangement.

Cancellations and credit balances

Cancel any time from the dashboard. The current billing period runs to its end (you keep access until then) and you're not charged again. Subscription tiers are non-refundable, but any top-up credit on the account stays there indefinitely — when you come back, it's still waiting.

Do I need a separate Patchworks subscription?

Yes. PatchBuddy uses your existing Patchworks API credentials to build flows in your Patchworks tenant. You'll need an active Patchworks subscription on your end.

What if I just want to try one build before committing?

The case-studies page on this site (linked from the header) is the closest thing — real builds, verbatim chats, the time and cost on each. We don't run a free-credit trial because £10 of model time can ship a working integration; the case studies pull back the curtain instead.

Skip the canvas.

Sign up, plug in your Patchworks credentials, and ship your first flow by chat. Same evening.