Verified AI Benchmarks • Updated September 2026
RankLLMs Authors & AI Researchers
Meet the researcher, benchmark analyst, and software engineer writing the technical LLM comparisons and reviews on RankLLMs.

Founder & Lead AI Benchmark Researcher
Lucky Yaduvanshi
38 articles published • 85 models tracked • writes every verdict on this site
Software engineer and AI performance researcher focusing on LLM latency, cost efficiency, SWE-bench coding benchmarks, and autonomous agent evaluation.
Why one author
RankLLMs is deliberately a single-reviewer publication. Every benchmark number, price, and verdict on this site passes one desk before publication, which means one person is accountable when something is wrong - and corrections happen in public, dated, on the editorial policy's terms. It also means no committee blending opinions into verdicts that satisfy nobody: when a comparison picks a winner, one researcher is putting their name on it.
The trade-off is throughput, not rigor. Instead of scaling the byline count, the work scales through systems: a documented scoring methodology, dated verification on every scorecard, and a public dataset behind the leaderboard that anyone can check line by line.
Editorial standards
- Dated verification
- Every scorecard carries the date its data was last checked; historical results keep the model version they belong to.
- Named sources
- Claims trace to official model cards, vendor pricing pages, independent benchmark organizations, or documented in-house test runs.
- Disclosed costs
- Cost-per-task math always states its token workload assumptions so you can adjust them to your own usage.
- Independence
- No vendor previews, influences, or purchases a ranking. Funding is disclosed on the About page.
- Public corrections
- Errors are fixed visibly and dated - never silently overwritten. Report anything via the contact page.
- Real testing
- CLI agent reviews come from documented runs on real repositories, not press-release summaries.
Most recent publication: GLM-5.3-FlashX: 200 Tokens/sec Throughput vs 2.5x Price Analysis. Want to propose a correction or a comparison? Contact the team.