Prediction: Every Future LLM Will Ship With a Native 'Jev Mode'
multimodalart · x · 2026-09-18
A multimodal-focused practitioner speculates that jev will be folded into all LLMs: new models will come with a native "jev mode" baked in, much like MTP-style acceleration is now standard. He says that if he were a lab training a new LLM, he'd try to be first to ship such a hybrid model. A personal prediction about where model architecture is heading.
Related event: Developers Predict 'Jev Mode' Will Become Native in All LLMs(2 posts)→
More from Models
- Qwen shares pricing page for the new Qwen3.8-Omni-Flash omni-modal model — Alibaba_Qwen · 2026-09-18
- Alibaba launches Qwen3.8-Omni-Flash, its first omni-modal model built for agentic workflows — Alibaba_Qwen · 2026-09-18
- Bonsai quant hits 50 tok/s at 128k context on a 24GB card, letting users run two sessions at once — julianharris · 2026-09-18
- Encoders and decoders are the same thing, and decoders have been doing classification for years — HanchungLee · 2026-09-18
- Sakana AI Launches Fugu Max, a Multi-Agent Orchestrator Routing Tasks to Leanest Capable Models — tkasasagi · 2026-09-18
- Why don't LLM labs compete on personality? Users value it over raw capability — dioscuri · 2026-09-18