- Published on
A Free, Ad-Free AI Office Suite for macOS and Windows with Docs, Sheets, Slides, PDF
- Authors
- Name
- DeepHorus
Genspark has open-sourced GenOffice, an AI-native office suite for macOS and Windows, under Apache License 2.0. The suite includes Docs, Sheets, Slides, and a PDF tool, built as five Electron apps sharing one engine layer. AI editing is a first-class workflow, not a bolted-on chat box. Signed installers ship for Apple Silicon and Windows x64 at version 0.4.110.
GenOffice is free and ad-free for everyone. The team’s X post says the Alpha was built by one engineer in one week, costing roughly $10,000 in tokens. AI features consume Genspark credits, and feedback goes through a GenTeam group chat with 1,000+ credits for active contributors.
Deployment is possible today via signed binaries or source builds with Node 20+ and Rust. But the AI path requires a Genspark account because model calls are proxied service-side. The project is explicitly labeled an Alpha.
The license benefits startups and SMBs most: no fees, no watermarks, full Office file fidelity. Mid-market teams can fork the code under Apache-2.0, though the ee/ directory is reserved for future enterprise modules, and Genspark trademarks cannot be reused. Regulated enterprises should wait due to alpha status and cloud-routed AI.
The technical core ensures byte-preserving round trips. GenOffice hashes the original .docx and never modifies it; only dirty blocks are converted to OOXML and spliced back on save. Untouched blocks keep their original bytes, so Word layout is never broken. Sheets and Slides follow the same patch-based philosophy.
Sheets builds on the open-source Univer core with a Rust sidecar for XLSX. Slides uses an in-house pptx engine, and PDF is built on pdf.js and pdf-lib. Security is detailed: every window uses sandboxing, and IPC payloads are schema-checked.
Looking ahead, this open-source release could challenge incumbent office suites by offering a free, AI-native alternative with strong file fidelity. Yet the cloud-dependent AI and alpha maturity mean full enterprise adoption will take time. Continued feedback from the open-source community may accelerate development and stability.
