Edge0-145B: Frontier AI capabilities on your own computer
13 November 2026 · filed under f980cc1f1e21
Edge0.ai announced Edge0-145B on November 12, describing it as a 145-billion-parameter model intended to deliver “frontier AI capabilities” on consumer-grade hardware, according to the company’s blog post published that day. The announcement was shared to Hacker News, where it drew discussion from the technical community.
The post, titled “Edge0-145B: Frontier AI Capabilities on Your Own Computer,” frames the release as a step toward running large-scale models locally rather than through cloud-based inference services. Beyond the company’s own characterization, the available material does not specify benchmark comparisons, hardware requirements, licensing terms, or release timeline in detail.
The claim’s significance rests on the gap between model size and deployment location. Models in the 145-billion-parameter range have historically required data-center-class infrastructure to run at usable speeds, which has concentrated access to such capabilities among organizations with substantial compute budgets. A genuinely local-capable model at this scale would alter that arrangement, touching on questions of data privacy, who can afford to run frontier-tier systems, and the broader economics of AI inference.
No independent verification, third-party benchmarking, or technical review of the claims was available in the cited material at time of writing. The sole source remains the company’s own announcement and the ensuing Hacker News thread.