We are building the internet's most private, most comprehensive search and technology platform — without collecting a single byte of your personal data.
To make the internet safer and more private for everyone. We believe that great technology should not require you to surrender your identity, your habits, or your future to a corporation that profits from your data.
OBLIVION was built on a simple conviction: privacy is not a premium feature. It is a fundamental right. Every product we ship, every line of code we write, is guided by one question — "Can we build this without knowing anything about the user?" Almost always, the answer is yes.
OBLIVION powers a complete, privacy-first technology stack — anchored by the world's most comprehensive private search engine.
OblivionSearch simultaneously queries 246 independent search engines and aggregates the best results. No single algorithm controls what you see — you get the full breadth of the open web.
Our proprietary Scam Shield layer detects and removes fraudulent websites, phishing pages, and fake storefronts before results ever reach you. It runs on every search, automatically.
OBLIVION AI generates answers using models that run entirely on our own servers. Your queries never touch OpenAI, Google, Anthropic, or any third-party AI cloud.
No cookies. No fingerprinting. No user accounts required. No search history stored. No behavioral profiles built. Your searches are yours alone — we never see them.
OBLIVION serves users across more than 30 languages with full localization — making private search accessible to users everywhere, not just English speakers.
Beyond search: OBLIVION Mail, Pages, Analytics, Support, AI, Business, Tools, and Developer APIs — an entire privacy-respecting platform under one roof.
Most search engines rely on a single index controlled by a single company. OBLIVION takes a fundamentally different approach — federated search intelligence combined with on-device AI and multi-layer verification.
Every search simultaneously queries up to 246 independent search engines — Google, Bing, DuckDuckGo, Brave, Mojeek, and 241 more. Results from all sources are collected in under 2 seconds. No single company controls what you find.
Results are ranked using a consensus algorithm inspired by academic research (BM25, HITS, and federated rank aggregation). If multiple independent engines agree a result is relevant, it rises to the top — eliminating single-source bias.
Every result passes through our proprietary safety layer. Domain age, SSL validity, known phishing databases, content fingerprinting, and behavioural signals are scored in real-time. Dangerous results are flagged or removed before you see them.
15 AI models running on our own Tesla T4 GPU generate instant answers, knowledge panels, and search suggestions. Models include Qwen3 (235B parameters), LLaMA 3.3 (70B), DeepSeek-R1, and Phi-4. Your queries never leave our servers.
21 specialised search verticals — Web, Images, Videos, News, Academic, Finance, Weather, Nutrition, Medical, Legal, Patents, Local, Music, Books, Code, and more — each powered by dedicated APIs and curated data sources.
The entire pipeline runs without storing your IP address, setting tracking cookies, building user profiles, or sharing query data with third parties. Privacy is not a feature we bolt on — it is the architecture itself.
OBLIVION runs on dedicated hardware — not rented cloud instances that can be subpoenaed or shut down. Our infrastructure is designed for sovereignty, speed, and resilience.
64 CPU cores, 393GB RAM, Tesla T4 GPU, 7.7TB storage across 5 dedicated drives. We own our hardware — no AWS, no Google Cloud, no Azure. Your data never touches Big Tech infrastructure.
PostgreSQL, Elasticsearch, ClickHouse, Qdrant (vector search), Meilisearch — each purpose-built for specific workloads. From full-text search to AI embeddings to real-time analytics.
Docker containers running SearXNG, Heritrix web crawler, LibreTranslate, Whisper (speech-to-text), IPFS, and dozens more open-source tools. All self-hosted, all auditable.
HTTPS, Tor (.onion hidden service), IPFS (decentralised), Gemini protocol, Yggdrasil mesh network, GUN.js peer-to-peer, and Blockstream Satellite. Censorship-resistant by design.
OBLIVION was founded in 2026 by Vasile Gangal with a conviction that the internet had taken a wrong turn.
For two decades, the dominant model of the web has been surveillance capitalism: give users free services in exchange for unprecedented access to their most intimate data — their searches, their locations, their relationships, their fears, their desires. That data is then sold to advertisers, governments, and data brokers.
Vasile founded Oblivion Technologies LLC to build a different model. One where the technology is just as powerful, the search results are just as comprehensive — but the user walks away having given up nothing. No profile. No cookie. No record. Oblivion of your data, by design.
Incorporated in Wyoming, USA — one of the most privacy-protective legal jurisdictions in the country — OBLIVION operates transparently, with a registered EIN, DMCA agent, and public Wikidata entry. We are a real company with a real address, built on the conviction that private technology is the future.
The scale of what OBLIVION offers — built by a lean, focused team committed to doing more with less.
OBLIVION is built on a foundation of open, auditable, privacy-preserving technologies — no black boxes, no proprietary surveillance stacks.
The open-source meta-search engine that powers our 246-engine aggregation. Fully self-hosted, no telemetry, no external calls except to the search engines you query.
All 14 of our AI models run via Ollama on our own servers. Inference stays on-premises. Your prompts never leave our infrastructure, let alone reach a third-party cloud.
High-performance full-text indexing powers our Scam Shield database and structured search features — enabling sub-100ms filtered results at scale.
OblivionSearch is accessible as a Tor hidden service, providing a completely anonymous access path for users who need the strongest possible network-level privacy.
Decentralized content addressing via IPFS ensures that our platform remains accessible and censorship-resistant, even in restrictive network environments.
We build on open standards: REST APIs, standard HTTP, no vendor lock-in. Our developer platform lets anyone integrate OBLIVION into their own applications.
These are not aspirational guidelines. They are hard engineering constraints we have built into every layer of the platform.
We set no tracking cookies. Session state is handled without identifying you across visits.
No user-level analytics. No behavioral profiling. No retargeting pixels. No third-party scripts that phone home.
We do not sell, license, rent, or share user data with any third party. Our business model does not depend on monetizing your attention.
Our AI runs on our servers, not OpenAI's or Google's. Your prompts are processed in-house and discarded — never stored, never trained on.
The core search experience requires zero registration. No email, no phone, no identity — just search.
Incorporated in Wyoming — a state with strong LLC privacy protections and no state income tax — keeping our legal structure clean and transparent.
A lean team with a sharp focus: build the most private, most useful search platform on earth.
Vasile founded Oblivion Technologies LLC in 2026 to address one of the internet's most persistent failures: the surveillance business model that trades user privacy for free services. He oversees product, engineering, and strategy — guided by the principle that powerful technology and genuine privacy are not in conflict.
Oblivion Technologies LLC is a registered legal entity with full public transparency about who we are and where we operate.
OBLIVION operates transparently within established legal and institutional frameworks.
Applied to the NLnet Foundation — Europe's leading funder of open internet infrastructure — for support in building privacy-preserving search technology accessible to all.
Registered DMCA agent with the U.S. Copyright Office (Agent #DMCA-1070891), fulfilling our legal obligations and providing a clear channel for intellectual property concerns.
Oblivion Technologies LLC and OblivionSearch are listed on Wikidata, the open knowledge base, establishing our public presence in the global linked-data ecosystem.
Incorporated and in good standing in Wyoming — a jurisdiction recognized for strong privacy protections, straightforward business registration, and transparent legal structure.
Join users who have switched to a search engine that treats them as people, not data points.