PayPerQ lets users access hundreds of AI models with Monero and other cryptos through a strict pay-per-query model that requires no subscriptions or accounts.
PayPerQ (PPQ.ai) operates as a pay-per-use aggregator that provides access to hundreds of text, image, video, and voice models from OpenAI, Anthropic, Google, and other providers. The service runs through a web app, an OpenAI-compatible API, and additional integrations, with no subscriptions or mandatory registration required.
Founded in 2024 by Matt Ahlborg and CTO Rafael Valverde, the platform completed a $500K early-stage VC round on August 12, 2025. It positions itself for users who want to pay only for actual queries or generations rather than recurring commitments.
The service appeals most to privacy-minded developers and individuals who prefer crypto payments, including Monero (XMR), Bitcoin, and Litecoin, alongside credit cards. Conversations store locally by default, and users can top up balances with a $0.10 minimum while choosing optional accounts or passkeys. This combination of flexible crypto funding and minimal data retention suits those who seek model variety without long-term data exposure or fixed fees.
Users begin by adding funds to a balance that starts at the $0.10 minimum top-up. The platform then deducts the precise cost of each generation or token batch directly from that balance. Average query cost sits at 1.5 cents according to the pricing page accessed in October 2026, with homepage copy and secondary references consistently placing typical usage in the 1.5–2 cent range.
These per-use charges produce an average monthly spend of $4 for active web-app users. Credit-card payments are supported alongside several cryptocurrencies, explicitly including Monero. Once topped up, the balance remains available until exhausted, with no automatic renewals or expiration dates tied to calendar months.
The model therefore differs sharply from subscription services that bill a fixed amount regardless of activity. Occasional users avoid paying for idle periods, while heavy users pay only for the volume they actually consume. Balance visibility and transaction history appear in the interface so users can track remaining credit before each session.
PayPerQ stores conversations locally by default, so nothing leaves the device unless the user deliberately enables optional accounts or passkeys. Passkeys add client-side encryption to chats that remain on the user’s hardware even when synced across sessions.
The Enclave Router, introduced in late September 2026, routes prompts through AWS Nitro enclaves. Prompts are encrypted on the client before transmission; the service receives only billing metadata. TEE models operate under the same model, keeping raw inputs inaccessible to the platform operator.
Monero is listed among supported payment methods alongside Bitcoin and Litecoin. Users can top up balances with XMR without creating an account, placing them in the Anon tier. Zero Data Retention and TEE tiers add further controls over what metadata is retained.
Practical OPSEC steps differ by tier. In the Anon tier, fund a local Monero wallet from a non-KYC source, connect through Tor or a trusted VPN, and never reuse addresses. For Zero Data Retention, still avoid logging in from identifiable networks and clear browser storage after each session. TEE users should verify enclave attestations where offered and keep the client device hardened, since endpoint compromise can bypass enclave protections.
PayPerQ lists per-million-token rates for aggregated models on its pricing page as of October 2026. Claude Haiku 4.5 carries input costs of $1.05 and output costs of $5.28. Claude Sonnet 5.5 is listed at $2.11 input and $10.55 output. GPT-5.4 Nano shows the lowest rates at $0.21 input and $1.32 output. The platform also supplies median costs across its catalog along with prompts-per-dollar figures for quick budgeting.
| Model | Input per 1M tokens | Output per 1M tokens |
|---|---|---|
| Claude Haiku 4.5 | $1.05 | $5.28 |
| Claude Sonnet 5.5 | $2.11 | $10.55 |
| GPT-5.4 Nano | $0.21 | $1.32 |
AutoClaw routing selects lower-cost models dynamically and can deliver up to 70% savings versus manual selection. Users can review the full rate sheet to compare cache rates and other options before running queries.
PayPerQ introduced Multi-Model Chat around 19 August 2026, letting users run up to four models in parallel for comparison and synthesis within a single session. The Memory feature arrived on 6 August 2026 and stores optional, user-controlled context across separate chats. Fireworks AI became the default open-source routing layer in July 2026, while expanded uncensored model options and private TEE tiers rolled out earlier in the year alongside passkey-encrypted storage.
The Enclave Router, released as of late September 2026, routes prompts through AWS Nitro enclaves with client-side encryption so that only billing metadata reaches PayPerQ servers. AutoClaw smart routing, promoted in February 2026, claims up to 70 percent savings by selecting cheaper models automatically.
These additions improve flexibility and privacy controls, yet the service still publishes no public user counts, total queries, or aggregate statistics. Without those metrics, traders and researchers cannot independently verify claimed cost savings or adoption trends beyond the per-query pricing already listed on the site.
You can begin immediately by topping up a balance with a $0.10 minimum via credit card or supported cryptocurrencies including XMR. No mandatory registration is required for basic use.
Select Monero at checkout on the payment page, send the exact amount shown to the generated address, and the balance updates once the transaction confirms on the Monero network.
By default, conversations store locally in your browser. Optional account or passkey features allow encrypted cloud sync, while TEE models keep prompts encrypted client-side with only billing metadata visible.
Hundreds of text, image, video, and voice models from OpenAI, Anthropic, Google, Fireworks AI, and others are available through the web app and OpenAI-compatible API.
AutoClaw smart routing selects cheaper equivalent models for up to 70% savings. You can also set routing preferences and monitor per-query costs, which average around 1.5 to 2 cents.
Basic pay-per-query access works without accounts. Passkeys or accounts unlock optional encrypted chat storage and cross-device memory features that remain user-controlled.
