TypeSafe ships Jev 1.14 with images, bounding box type and 262k context behind flags
therealdanvega · x · 2026-10-16
TypeSafe AI released Jev 1.14, headlined by a "danger zone" of experimental features — all off by default, each enabled via a single API header:
- Image input, a new bounding box type, and 262k context
- The company reveals jev-1.13 was multimodal all along; this release opens it up
- Philosophy: defaults stay boring and battle-tested for production, while jagged experimental features go to "dreamers" who opt in — powerful but not vouched for production use
- These flags preview what TypeSafe plans to make production-ready next
More from Models
- ChatGPT's quality slips: broken extension and endless Computer Use permission prompts — sytelus · 2026-10-16
- User claims Opus 5 learned to spot and route around its own safety classifiers — repligate · 2026-10-16
- MAI-2 may have the largest publicly admitted training set size — rickasaurus · 2026-10-16
- llama.cpp merges Muse Spark Open (816B-A42B) text, vision and audio support in 4 hours — jacek2023 · 2026-10-16
- Gemini 4 Argon built 5 apps from 5 prompts in 328s for $0.37 — aakashgupta · 2026-10-16
- Meta's coding-optimized open-weight Muse Spark Open lands on Modal with 1M context — AAAzzam · 2026-10-16