Perplexity CTO's AutoJev-27B Takes Open-Source Lead on Decision Index, Trained for Just $3k
antoine_chaffin · x · 2026-09-24
The Decision Index leaderboard updated to 0.2: improved formula, +29 Jev-like models, +21 benchmarks.
- AutoJev-27B, trained by Perplexity CTO Denis Yarats, took the open-source lead, trailing Jev by just 0.8 points
- Yarats revealed the training cost only $3,000, with more iterations running to hopefully surpass the top model
The index helps users find the best decision model at every size, speed, and use case. Near-SOTA at $3k is a striking signal for small-budget training.
Related event: Decision Index 0.2 released; Perplexity CTO's model tops open ranking(2 posts)→
More from Models
- User claims MiniMax H3 was trained on The Will Stancil Show after model adds unprompted whistle — Kyrannio · 2026-09-24
- Xiaomi's MiMo V2.6 Pro builds a habit tracker in 62 seconds for under 5 cents — socialwithaayan · 2026-09-24
- Gemini 4 Is Almost Ready, Says New Google DeepMind Chief Koray Kavukcuoglu — The Verge AI · 2026-09-24
- New chat, no refusal: user shows ChatGPT's self-image prompt only blocked in original thread — Todesluke · 2026-09-24
- Open Replications Miss the Secret Sauce: HF Collection Curates Datasets for Jev-Style Models — vanstriendaniel · 2026-09-24
- GPT-6 Astra Fixes a UI That Even Opus 5.5 Couldn't Crack — haltakov · 2026-09-24