Closed models look highly reproducible, while open weights still lag badly
zakkohane · x · 2026-07-23
The follow-up result is blunt: reproducibility looks solved for the closed models, but not for open weights.
- Closed/commercial models cluster high: GPT-5.2 0.94, Claude Opus 4.8 0.92, Grok 4.5 0.89, Gemini 2.5 Pro 0.85.
- Open-weight models remain much less stable: Kimi K2-Thinking 0.28, Llama 3.3 70B 0.24, GLM-4.7 0.14.
- The chart marks 0.8 as the “highly reproducible” line; the open-weight models sit well below it.
Related event: Clinical Triage Retest: Closed-Source Models Far Outperform Open-Source(6 posts)→
More from Models
- Astra Scores 83% on GauntletBench, First Computer-Use Agent to Beat Human Baseline — ducha_aiki · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- DeepSeek V4 Pro API to continue after Sept 2026, billing unchanged — teortaxesTex · 2026-09-11
- DeepSeek V4.1 Flash Hits 98% of GPT-6 Astra's Score at 1.4% of the Cost in Third-Party Benchmark — ayushtweetshere · 2026-09-11
- TheZvi Polls: Has Your Coding Model Choice Changed Since Fable 5.1 and Astra? — TheZvi · 2026-09-11