Kimi-K3 Preliminary ECI Evaluation Score Leaks
scaling01 · x · 2026-07-19
Kimi-K3 achieved a preliminary ECI (Effective Coding Intelligence) score of 155.53 (90% confidence interval: 153.87–158.21). This score slightly edges out Opus 4.6 (155.31) and GPT-5.3-Codex (155.36). Since this is still preliminary data, the original author notes that the confidence interval should reflect this uncertainty.
Related event: Kimi-K3 Preliminary ECI Score Surpasses Top Models(4 posts)→
More from Models
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- GPT-5.6 writes well but is instantly forgettable, user complains — BasedRaddka · 2026-09-11
- Opus Refuses Protein Research Codebase Over 'Safety' Concerns, Dev Considers Rolling His Own — josephdviviano · 2026-09-11