Eli Chen
Editor-in-chief · China AI Bench
Independent developer and editor of China AI Bench. I test Chinese AI models hands-on with my own API keys: practical tasks, fixed sampling parameters, versions snapshotted, and I publish the task sets and raw logs alongside the verdicts. Two models are in-house so far (DeepSeek V4 Pro and V4 Flash, August 2026); the rest of the board is compiled from vendor docs, independent evals, and community tests, with every number linked to its source. English first, quarterly retests of the in-house pair, 48-hour spot tests after major releases.
Articles
News Desk and Benchmarks Desk are organizational bylines (collective editorial desks — see About); the editor is Eli Chen, a real person with a verifiable GitHub profile. Future individual contributors reuse this page pattern.