Sigma Launches the First Private AI Browser With Cloudless LLM

SAN FRANCISCO, CALIFORNIA / 18 December, 2025 / Sigma Browser has announced the launch of Sigma Eclipse, the world’s first AI-powered browser built on a fully local, cloudless large language model architecture.

In contrast to existing AI browsers that rely on remote servers to process user prompts and automate tasks, Sigma Eclipse executes all AI operations directly on the user’s device. This local-first design ensures that browsing data, credentials, personal context, and task inputs remain entirely on-device, without being transmitted to external systems. Users are therefore able to take advantage of AI-powered workflows without sacrificing control over sensitive information.

Alex Shapiro, CSO of Sigma Browser:“At some point, privacy stopped being a setting and became a marketing term, with Sigma Eclipse, we wanted to make privacy verifiable rather than promised. When the AI runs on your own machine and doesn’t require an internet connection, users can see for themselves that their data stays exactly where it belongs”.

Sigma Eclipse’s execution model enables AI chat to function even offline, reinforcing transparency around data handling. The browser is built on an open-source codebase. According to the company, this approach is intended to eliminate hidden data collection, opaque processes, and undisclosed access points typical of closed-source AI systems.

Alongside its local AI foundation, Sigma Eclipse introduces an uncensored language model that does not impose ideological constraints or content filtering on prompts and responses. Sigma states that this decision is rooted in the belief that users should maintain full autonomy over their interactions with AI, without restrictions on subject matter or viewpoint.

The release also adds support for local PDF analysis, enabling users to review and work with documents locally. Files are processed entirely on the user’s machine and are never uploaded to third-party servers, making the feature suitable for confidential materials such as legal documents, financial records, or sensitive professional files.

Nick Trenkler, cofounder of Sigma Browser. “AI has become incredibly powerful, but it has also become centralized and expensive. We believe users shouldn’t have to trade privacy or pay ongoing cloud costs to access advanced AI. By keeping everything local, we’re able to offer a private LLM that’s free to use and free from surveillance”.

Because Sigma Eclipse does not rely on cloud-based processing, the company reports that it avoids the operational costs typically associated with AI infrastructure. As a result, the core local LLM functionality is offered to users for free, without requiring a subscription.

This release is a step toward a more transparent and user-controlled AI ecosystem, where performance, privacy, and accessibility are not mutually exclusive. Sigma describes this release as a move toward a more open and user-centric AI landscape, where privacy, performance, and accessibility coexist. The company believes that local-first AI systems present a credible alternative to cloud-dependent models, especially as concerns over data ownership, security, and long-term costs continue to intensify.

‍