Frontier models improve on earnings-direction benchmarks, but open models still lag
dougclinton · x · 2026-07-21
- Intelligent Alpha’s Q2 earnings benchmark shows newer frontier models generally improved versus their predecessors on predicting whether revenue/earnings expectations would move up, down, or stay flat.
- Fable 5, Grok 4.5, Kimi K3, MiniMax 3, and GLM 5.2 all beat their prior versions on direction accuracy.
- GPT 5.6 was the only newer model that underperformed its predecessor on this benchmark, though it still beat all the other models.
- The same broad pattern held on the harder task of predicting the magnitude of revenue/earnings changes.
- The author concludes there is still a clear gap between frontier closed models and open models for investment decision-making.
More from Venture
- Polymarket prices a 17% chance the AI bubble bursts by year-end — Polymarket · 2026-07-21
- Déjà View looks up earlier startups for any idea and how they ended — Sea-Assignment6371 · 2026-07-21
- Open-source AI could capture enterprise spending as closed-model pricing keeps eroding — bigdata · 2026-07-21
- TSMC’s 3nm utilization reportedly tops 120% as AI demand drives a $190B capex cycle — tengyanAI · 2026-07-21
- Chinese AI startups rush to raise capital as U.S. rivals still pull in more cash — KateClarkTweets · 2026-07-21
- Cordant exits stealth with an $8M seed round for payments visibility tools — LexSokolin · 2026-07-21