Mullvad Leta

Mullvad Leta is a *privacy-by-proxy search front-end* bundled with a Mullvad VPN subscription — it runs your query against a major upstream engine through Mullvad's own infrastructure, stripping the identifying signals before they ever reach the source.

Background. Launched in 2023 by Mullvad, the well-regarded no-logs Swedish VPN, Leta exists to close an obvious gap: you can hide your traffic behind a VPN and still leak your entire search history to Google the moment you type a query. Leta proxies searches (Google and, optionally, Brave results) through Mullvad's servers and serves *cached results*, so the upstream engine sees Mullvad's infrastructure rather than you. It is gated to active Mullvad subscribers. Coming from an operator with Mullvad's audited no-logs track record — and one that famously accepts cash and assigns no personal account identity — is why it earns an A in /search despite being a proxy rather than an index.

What you trust. You're trusting Mullvad, and Mullvad has earned more of that trust than almost any operator in the space: a *no-logs* posture validated by third-party audits, an account model that is just a random number (no email, no name), and a payment story that includes cash and crypto. Leta inherits that institutional credibility — the same company that won't log your VPN traffic is the one proxying your searches. Because results are *cached and proxied*, the upstream engine doesn't see your IP, your VPN exit, or a per-user identity; it sees Mullvad fetching a query. The trust is concentrated in one well-audited operator rather than scattered across an ad-tech supply chain.

Operational specs. Access requires an *active Mullvad subscription* (the same random account number that powers the VPN) and is capped at *100 queries per day per account* — a deliberate anti-abuse limit, not a paywall upsell. Leta proxies an upstream index (Google by default, with a Brave option) and returns cached results, so you get familiar mainstream-quality answers without the mainstream tracking. There's no separate signup, no email, no additional fee beyond the VPN subscription. It's a lean, fast, text-first interface in keeping with Mullvad's minimalist design ethos.

Philosophy. Mullvad's whole brand is *minimal identity and minimal logs* — they reduced the account to a number and the data trail to as close to nothing as a functioning service allows. Leta is that philosophy applied to search: rather than build a full independent index (enormously expensive), Mullvad inserts itself as a privacy buffer between you and an existing high-quality index, absorbing the identifying layer. It's an honest, scoped solution — "we can't out-crawl Google, but we can stop Google from seeing you" — and the per-day cap reflects a refusal to turn it into an ad-funded scale product.

Grade rationale. A in /search. The grade reflects the operator's audited no-logs reputation, the identity-stripping proxy design, cached-result delivery, and an account model (random number, cash/crypto payment) that's among the most privacy-respecting anywhere. It is explicitly *not* an independent index — that's the structural distinction from Brave Search — and the A is earned on trust-in-operator and proxy design rather than on owning the crawl.

Useful when. Use Leta when you're already a Mullvad subscriber and want mainstream (Google-grade) results without handing your query stream to Google — it's the natural default search for the privacy-minded Mullvad user. The 100/day cap is plenty for ordinary browsing; pair it with another engine for heavy research days. It's especially apt when you specifically want Google's result quality (not an independent index's) but none of Google's profiling.

Caveats. The hard gate is real: *no Mullvad subscription, no Leta* — it's not a standalone public engine. The *100-queries-per-day* limit will bite power users mid-research, so it's a complement, not always a sole engine. You are trusting a single operator (Mullvad) and, indirectly, the upstream index's result quality; the proxy hides *you* from Google but the results still originate from Google's ranking. And because it's a proxy, it offers no independent index resilience — if the upstream relationship changed, the product would have to adapt. None of this undercuts the A: for the specific job of "Google results without Google seeing me, from an operator with a proven no-logs record," Leta is close to best-in-class.