TypeSafe releases Jev 1.14 with experimental image input, box output and 262K context

1. 12:45 PM
   Vercel Developers@vercel_devOfficial

   Jev 1.14 from Typesafe AI is now available on Vercel AI Gateway

   AITypesafe AI's Jev 1.14, version typesafe-ai/jev-1.14.0, is now available through Vercel AI Gateway. The release adds typed questions about images and experimental long context support, and both features were already supported in Jev 1.13.

2. 1:45 AM
   MarkTechPostNews

   TypeSafe AI releases Jev 1.14 with images, box type and 262k context

   AITypeSafe AI has released Jev 1.14, a new version of its System One decision model, now live under the jev-preview alias. It adds three experimental features: image input, a box type that returns bounding boxes, and an opt-in 262k-token context, none of which TypeSafe says are production-ready. Pricing holds at $0.042 per million input tokens with free output, and the model is available only through TypeSafe's hosted API.

3. 1:36 AM
   Guizang@op7418X

   JEV 1.14.0 adds multimodal support, box output, and 262K context

   AIJEV released version 1.14.0, adding multimodal support that the source says the model already had but lacked in its first version. The update adds a box output type that returns object positions in an image and raises context to 262K tokens per request and answer. The source says multilingual ability also improved.

   

1. 10:56 PM
   Guizang@op7418X

   JEV 1.14.0 adds multimodal support and 262K context window

   AIJEV released version 1.14.0, adding multimodal support, which the official team says the underlying model already had but was not enabled in the first release. A new box output type lets the model return the position of a target within an image, and context per request and per answer rises to 262K.

   
2. 9:02 PM
   TypeSafe AI@typesafeaiOfficial

   jev-preview now points to jev-1.14.0 with multilingual gains

   AITypeSafe AI says the jev-preview alias now points to jev-1.14.0, with subtle improvements across the board. The company expects the largest gains in multilingual support and asks users to report where Jev still performs poorly.

3. 9:02 PM
   TypeSafe AI@typesafeaiOfficial

   TypeSafe AI adds a bounding box primitive to Jev 1.14

   AITypeSafe AI introduces a bounding box primitive in Jev 1.14 for locating objects at Jev's speed. The company says the primitive goes beyond decisions and that it carefully defined its System One models. Documentation is available at

   
4. 9:02 PM
   TypeSafe AI@typesafeaiOfficial

   TypeSafe AI releases Jev 1.14 with image, box, and longer-context previews

   AITypeSafe AI announces Jev 1.14, a new and improved model, released on 10/14 at 1:14pm PT. Experimental preview features include image input, a new box type, and longer context, which the company warns are not production-grade.

   
5. 8:16 PM
   OpenRouter@OpenRouterOfficial

   TypeSafe publishes release notes for its jev 1.14 update

   AIOpenRouter links to TypeSafe's blog post with the full release notes for jev 1.14. The post itself contains no further details in the source text.

6. 8:16 PM
   OpenRouter@OpenRouterOfficial

   Jev 1.14 arrives on OpenRouter with multilingual gains and experimental vision

   AIOpenRouter says Jev 1.14 is now available, with multilingual performance as its biggest upgrade and experimental image input added. Per @typesafeai, the release also improves slightly across the board, while input pricing stays at $0.042 per million tokens and output is free.

7. 5:58 PM
   Yuchen Jin@Yuchenj_UWX

   Jev 1.14 adds image inputs, Box API, and longer context

   AIJev 1.14 is released with image inputs, a Box API, better output quality, and longer context, according to Yuchen Jin. Databricks says it is partnering with TypeSafe to bring the new Jev to its customers.

The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.