ownify.pricing

Pay for an agent that actually works —
and can prove what it did.

Plans from €12/month. 14-day free trial. Cancel any time. Every plan includes an isolated agent runtime and, when you activate it, a MolTrust-verifiable identity.

What can your agent do?

Here are real things people set their ownify agents up to do.

📅

Handle scheduling

Your agent reads your calendar, replies to meeting requests, proposes times, books rooms. Works with Google Calendar, Outlook, and any Cal.com-style booking link.

📧

Triage your inbox

Reads incoming emails, drafts replies for your review, flags phishing, archives noise. Never deletes without asking (on Balanced preset).

🧾

Process invoices

Extracts line items from supplier PDFs, cross-checks against purchase orders, queues approvals. Signed receipts mean "we've already paid this" is provable.

🔍

Research and summarize

Reads long reports, extracts the decision-relevant parts, cites sources. Your agent keeps a memory of what it's read so follow-up questions stay in context.

🤝

Negotiate with other agents

Your agent contacts a supplier's agent to request a quote. Signed identity = the supplier knows they're talking to "Harald's agent" not a random bot.

🛎

Customer support on your site

Answers customer questions 24/7 using your documentation. Escalates to a human when confidence is low. Every interaction is logged for review.

Plans

Prices in EUR. VAT added at checkout via Stripe Tax. Top-up credits at 1.47× Fireworks list price (EUR), 12-month expiry.

Solo
€12/mo
  • 1 agent
  • €3.00 LLM per month
  • 10 GiB storage per agent
  • Audit trail + downloadable agent certificate
Start Solo
Duo
€22/mo
  • 2 agents
  • €6.00 LLM per month
  • 10 GiB storage per agent
  • Audit trail + downloadable agent certificate
Start Duo
Team
€32/mo
  • 3 agents
  • €9.00 LLM per month
  • 10 GiB storage per agent
  • Audit trail + downloadable agent certificate
Start Team
Enterprise
Custom
  • Unlimited
  • Custom LLM per month
  • Custom storage per agent
  • Audit trail + downloadable agent certificate
Contact sales
Pro
€99/mo
  • 10 agents
  • €30.00 LLM per month
  • 10 GiB storage per agent
  • Audit trail + downloadable agent certificate
Coming soon
Business
€279/mo
  • 30 agents
  • €90.00 LLM per month
  • 10 GiB storage per agent
  • Audit trail + downloadable agent certificate
Coming soon

Why run multiple agents instead of one big one?

Specialist agents beat a single generalist on both security and cost. This is why the plan ladder goes by agent count.

🛡 Security case

  • Smaller blast radius. If your email agent gets a prompt injection, only your inbox is at risk — not your calendar, bank, or customer database.
  • Least privilege. The invoice agent doesn't need HR data. Each agent only has access to what its job requires.
  • Separate identities. Each agent has its own DID + signing key. Compromising one doesn't compromise the others.
  • Independent audit trails. "The finance agent approved this" is cleaner evidence than "the mega-agent did it somewhere."

💰 Cost case

  • Shorter prompts = less spend. A specialist's system prompt is ~5× shorter than a generalist's. Tokens are the bill.
  • Cheaper models for simple tasks. Email triage can run on GPT-OSS 120B (€0.22/1M). Reasoning-heavy work picks DeepSeek V3.2 or GLM 5.1. Mix and match per job.
  • Parallel execution. Three agents working simultaneously on different tasks finish 3× faster.
  • Cached-prompt pricing. Stable system prompts hit the cached-input rate (e.g. €0.24/1M on Kimi K2.6 vs €1.40 uncached) — specialists benefit most.

Rule of thumb: 3 focused agents cost ~50-60% less than 1 generalist doing the same work.

Why only open-source, commercial-use models?

Your agent runs on weights you could, in principle, download and self-host. That’s a deliberate choice: no proprietary-API lock-in, no clauses that quietly restrict commercial use, no dependence on a single vendor’s pricing whim.

Every model in the catalog below ships under one of: Apache 2.0, Modified MIT, or DeepSeek License v1 — all permit commercial use at ownify’s scale. We don’t route to closed-model APIs, CC-BY-NC research checkpoints, or licences with acceptable-use clauses we can’t enforce.

The same stance applies to the platform itself — Kubernetes, Postgres, Zitadel, LiteLLM, cert-manager, and the rest run on commercial-use open-source licenses too. Full list: docs/open-source →

LLM pricing

Any of these models is available on every plan. Pick per agent from the agent’s Settings → Model tab. Your included LLM budget + wallet top-ups cover usage across all models. Pricing = Fireworks list × 1.47 treated 1:1 as EUR (covers platform + EUR/USD buffer).

Available modelsPrices in EUR · Fireworks USD list × 1.47
ModelLicenseInput €/1MCached €/1MOutput €/1MContext
klaw-auto
smart routing — picks the right model per request
—billed at the routed model’s rateup to 262k
GPT-OSS 120B
cheap + fast, OpenAI open-weights
Apache 2.0€0.22€0.01€0.88131k
DeepSeek V3.2
premium reasoning
DeepSeek License v1€0.82€0.41€2.47163k
Kimi K2.5
agent workhorse
Modified MIT€0.88€0.15€4.41262k
Kimi K2.6
newest agent model
Modified MIT€1.40€0.24€5.88262k
GLM 5.1
balanced reasoning
Apache 2.0€2.06€0.38€6.47202k

All models are commercial-use open-source, routed through ownify's LiteLLM proxy to Fireworks.ai (US, Zero Data Retention). The 47% markup covers the ownify platform + EUR/USD currency buffer. New agents default to klaw-auto (smart routing). Pin a single model any time in the agent’s Settings → Model tab.

Every plan includes

Isolated agent pod

Your own Kubernetes pod + storage. Not shared with other customers.

5 open-source LLMs

GPT-OSS 120B, DeepSeek V3.2, Kimi K2.5 + K2.6, GLM 5.1 — all commercial-use licensed. Transparent Fireworks pricing × 1.47.

Verified identity

Your agent gets a MolTrust DID, can be verified by any counterparty.

Signed action receipts

Every agent-to-agent message gets a signed "Interaction Proof Record" stored in our EU database.

EU-hosted

App + storage in OVH Germany. LLM inference via Fireworks.ai (US, Zero Data Retention).

14-day free trial

No card needed to trial. One agent, €0.50 LLM budget, full functionality.

Where does all this actually live?

The technical labels further down (audit retention, L2 anchor, key custody) can be confusing. Here's what they mean in plain terms.

Your conversations, documents, agent memoryownify-managed storage in OVH Germany (Frankfurt)
Encrypted at rest, persists across agent restarts. Only your ownify login can access it.
Agent certificate + event logOur Postgres database in OVH Germany
The registration credential (a signed W3C Verifiable Credential from MolTrust) and your agent's state-change log. Viewable and downloadable from your dashboard audit trail. Kept permanently while your account is active.
Per-interaction signed receiptsOur Postgres database in OVH Germany (+ Base L2 hash on Pro+ plans)
Generated when your agent interacts with ANOTHER MolTrust-enabled agent (A2A call, delegated action). Both sides sign a receipt of what was agreed. Kept 30 days (Solo) to 3 years (Business). For solo-agent use there's no counterparty to sign with, so these stay empty until the agent network around you also adopts MolTrust.
L2 anchor (blockchain proof)Base L2 — a public blockchain owned by nobody
On Pro+ plans, a cryptographic fingerprint of important receipts is written to Base L2. Nobody — including us — can alter or delete it afterwards. The raw content stays private; only a tiny hash goes on-chain.
Signing key (your agent's private key)Solo/Duo/Team: securely stored inside ownify\u2019s cluster. Pro/Business: your own cloud key vault (AWS KMS, GCP KMS, Azure Key Vault). Enterprise: hardware security module.
Your agent uses this key to sign its outbound messages. On lower plans, ownify holds it on your behalf (convenient). On higher plans, you hold it (you stay in full control).
Your payment methodStripe (EU-routed)
Card data never touches ownify. Stripe handles all PCI compliance and VAT.
LLM inferenceFireworks.ai (US) with Zero Data Retention
Prompts and completions flow to Fireworks for processing, then are NOT stored. Not used for training. Commercial-use open-source models only — Apache 2.0, Modified MIT, or DeepSeek License v1. No proprietary closed-model APIs, no CC-BY-NC weights, no non-commercial research checkpoints.
Show the full feature-by-feature comparison
Solo
€12/mo
What does each feature mean? →Start Solo
Duo
€22/mo
What does each feature mean? →Start Duo
Team
€32/mo
What does each feature mean? →Start Team
Coming soon
Pro
€99/mo
What does each feature mean? →Contact sales
Coming soon
Business
€279/mo
What does each feature mean? →Contact sales
Enterprise
Custom
What does each feature mean? →Contact sales

Want to see exactly where every euro goes?

We publish the unit-economics formula and plug it with our real production numbers — including a side-by-side of yesterday's margin shape vs today's vs the v3 forecast.

The math behind the pricing →

Want to understand the trust layer in detail?

We built a page explaining how ownify + MolTrust work — with diagrams, scenarios, and what happens step-by-step when your agent talks to another agent.

How the trust layer works →