Free
€0/month
2 million processed input tokens each month
Free
- Email verification required
- Claude and OpenAI gateways
- Dashboard and request logs
MONTHLY PRICING
Low-cost plans for every UVOX engine mode.
Choose a monthly allowance for successful processed customer input tokens. Auto, Maximum Savings, Zero-Loss Cache and Disabled use the same UVOX plan accounting. Failed provider requests deduct zero. OpenAI and Anthropic charges remain separate.
Starter
€9/month
50 million processed input tokens
€0.18 per million
- Production API access
- Usage and savings analytics
- Failed requests deduct zero
Growth
€29/month
250 million processed input tokens
€0.116 per million
- Team access
- AI-agent workload support
- Transparent usage evidence
Business
€99/month
1 billion processed input tokens
€0.099 per million
- Higher-volume production traffic
- Priority support
- Team and billing controls
Scale
€299/month
5 billion processed input tokens
€0.060 per million
- High-volume agent workloads
- Advanced onboarding
- Priority capacity planning
Enterprise
Custom
5B+ processed input tokens
From €0.04 per million
- Contracted volume and limits
- Enterprise onboarding
- Custom support and invoicing
Monthly allowance
Your plan includes the displayed number of processed input tokens for each billing month. Free accounts receive 2 million monthly tokens after email verification.
Charging rules
Successful provider requests deduct the original customer input-token count before any context optimisation. Failed provider requests deduct zero. Dashboard previews are not charged.
Provider billing
OpenAI and Anthropic charges are paid directly through the customer’s own provider account and are not included in UVOX plans.
Measurement notice: the historical Zero-Loss Cache audit completed 110 checks with zero failures and measured 38.52% gross provider savings across its mixed matrix. Combined Context + Cache is beta. Context reduction, cache discount and combined cost reduction use different baselines and are displayed separately. Results vary and are not guarantees.
Zero-risk verification is included in every plan
Every customer receives Zero-Loss Cache Mode, prompt-integrity hashes, provider/model route evidence, per-request cache economics and access to the independent verifier toolkit.
Zero-Loss Cache
No prompt rewriting, context deletion, semantic response reuse or provider/model switching.
Proof Lab
Compare direct cold, direct warm, UVOX cold and UVOX warm usage with your own keys.
Evidence ZIP
Download PDF, JSON, CSV, hashes and a signed certificate without keys or prompt content.
Independent verifier
Windows, Python, Node.js, Jupyter and cURL tools are available for download.
Claim basis: 64.7% was the median warm-cache cost reduction in a four-case Anthropic staging benchmark. Results vary and are not guaranteed.