JevBench v1.4.1 adds six systems; Jev-Ommi debuts at #7, top five unchanged
airesearch12 · x · 2026-09-24
Benchmark Heaven released JevBench v1.4.1, adding six systems to its Jev-class model leaderboard with the top five unchanged.
- Best newcomer: Jev-Omni, a multimodal Gemma-based model debuting at #7 with a composite score of 51.34
- Jev 1.13.0 holds #1 at 63.29
The underlying benchmark site also offers cost-vs-capability views, EU/sovereign hosting filters, and data-confidentiality options.
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