🔍 Read the full analysis: A Closer Look At Jev’s AI Vision With 30 Images on ThorstenMeyerAI.com
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
TypeSafe released Jev 1.14 on October 14, adding opt-in image input and a bounding-box response type. Thorsten Meyer tested Jev against 30 images from his website and reported strong performance on product photos, while saying the release remains too rough for production use without further evaluation.
Jev is built to answer structured questions rather than generate prose: users provide a state and typed questions, and the model responds with a probability, a choice from a list or a score. Before this release, those questions had to concern text. Jev 1.14 adds image support, which TypeSafe says also works with jev-1.13.0, and a new box type in version 1.14 that returns one bounding box and a confidence score for a specified target.
The image feature is behind an explicit request header, TypeSafe-Danger-Zone-Images: true. Images can be referenced in the request’s state, question instructions or criteria for answer options. TypeSafe also introduced an opt-in longer context limit: a separate header raises the per-request limit from 64,000 to 262,000 tokens, and the per-question limit from 32,000 to 262,000. The company advises against relying on the larger limit because reliability can fall on long inputs.
Meyer tested the model on 30 featured images from ThorstenMeyerAI.com: 10 product roundups, 15 news posts and five essays. He asked whether each image matched its own headline, a randomly selected other headline and an unrelated plant-care headline. For the product images, he reported that Jev 1.14 gave each image’s own headline scores of 0.65 to 0.89, while random headlines scored no higher than 0.18. It identified the featured product correctly in nine of 10 cases. These are results from a small, site-specific test, not an independent benchmark.
Jev 1.14 · images + box
Jev can see now. Here is what 30 of my images showed.
I asked about every featured image three times: does it fit its own headline, a random other post’s headline, an off-topic headline? Both Jev versions, about 60 requests.
Does this image fit the headline?
Median “yes” probability on Jev 1.14, by kind of post
- Product roundups10 images0.760.061.13: 0.79 / 0.05
- News posts15 images0.330.161.13: 0.51 / 0.25
- My essays5 images0.510.161.13: 0.62 / 0.33
Off-topic headline (“How to Repot a Monstera Plant”): never above 0.10 on either version, 60 of 60 rejected.
Why news is weak: almost all news images show the same scene, people at desks looking at monitors. Jev could not tell them apart, and it was right not to.
1.14 is stricter than 1.13
Same 30 images, same questions
| Result | 1.14 | 1.13 |
|---|---|---|
| Fits own headline (≥ 0.5) | 15/30 | 23/30 |
| Fits a wrong headline (≥ 0.5) | 1/30 | 4/30 |
| Own beats random other (AUC) | 0.91 | 0.89 |
| Names the product (10 options) | 9/10 | 10/10 |
1.14’s one product miss: mesh system read as “Wi-Fi 7 router”, confidence 0.60. Re-tune thresholds before jev-latest moves.
What shipped on Oct 14
- headerImagesin state, instructions or criteria · 1.13 + 1.14 ·
TypeSafe-Danger-Zone-Images - 1.14 onlyBox typeone bounding box + confidence. 6 of 6 clear targets boxed (0.66–0.90); an absent bicycle still got a box (0.03)
- headerLong context64k / 32k → 262k tokens, slower and less accurate
- opt-inHelp find weaknessesqueries used for testing, not training
now: jev-preview = 1.14~Oct 21: jev-latest → 1.14notice: 60 days (1.13), 30 days (1.14)
Source: own test on October 15, 2026. 30 featured images from ThorstenMeyerAI.com (10 product roundups, 15 news posts, 5 essays), JPEG at up to 1,568 px, jev-1.14.0 and jev-1.13.0 with the image header. Bars show medians. AUC = probability that an image’s own headline scores higher than a random other headline. Features, prices and limits: TypeSafe blog post “Jev 1.14” and docs.typesafe.ai, as of October 15, 2026.
Powered by Thorsten Meyer AI
Image Checks Move Into Structured AI
The release extends Jev from text-based classification to a task involving visual material, while keeping its core format intact: users ask defined questions and receive structured answers. That could let teams test whether an image suits a headline or locate a named object without asking a general-purpose model to write a description.
TypeSafe’s opt-in approach also signals that the company regards these features as experimental. Image support and longer context do not change default production behavior, and the company says it is not ready to vouch for them in production. That distinction matters for developers weighing whether to experiment with the feature or use it in workflows where mistakes carry costs.
As an affiliate, we earn on qualifying purchases.
Jev’s Experimental Release Controls
TypeSafe presents Jev as a model for decision tasks rather than open-ended writing. Users define the question and the allowed form of the answer, such as yes or no, a selection from a list, or a numerical score, with a confidence estimate. The October release adds images to that system rather than changing its basic purpose.
The company says Jev 1.14 is available as jev-preview. The default jev-latest alias is scheduled to move to the new version about a week after release, around October 21. TypeSafe also says it plans to give 60 days’ deprecation notice for version 1.13.0 and 30 days for version 1.14.0, with a formal policy still to come.
For image requests, TypeSafe documents support for PNG and JPEG stills between 256 and 1,568 pixels on each side. Each file can be up to 16 MiB, and a request can contain up to 32 MiB of images. The service does not automatically resize unsupported images; WebP and HEIC files need conversion. The company says image input has no separate surcharge and is billed as ordinary input tokens at Jev’s stated rate of $0.042 per million input tokens.
“There is “still way too much jaggedness for us to vouch for these for production use cases.””
— TypeSafe, in its Jev 1.14 release notes, as quoted by Thorsten Meyer
As an affiliate, we earn on qualifying purchases.
Small Test, No Production Endorsement
Meyer’s test covers only 30 images from one website, and the supplied account gives detailed results for the product-image subset but does not provide complete numerical outcomes for the 15 news images and five essays. The results cannot establish how the feature performs across other image types, publishers or real-world production workloads.
TypeSafe has not endorsed the new capabilities for production use, citing reliability concerns. Its release notes also describe internal intelligence estimates rather than public benchmark scores; Meyer says the comparisons involve open-weight models and may be biased toward Jev. The material does not provide independent evaluations or enough detail to assess those comparisons.
It is also unclear how the bounding-box feature performs across varied images and targets, or how often confidence scores reliably flag errors. Meyer’s product test suggests a confidence threshold might have identified the single Jev 1.14 miss, but that observation comes from only 10 product images and does not establish a dependable general rule.
As an affiliate, we earn on qualifying purchases.
Preview Rollout and Independent Tests
Jev 1.14 is available now through jev-preview, with TypeSafe planning to move jev-latest to it around October 21. Developers can test image input and bounding boxes with the required opt-in headers, while the default behavior remains unchanged until then.
TypeSafe says it will provide deprecation notice for the older model versions, but a formal policy is still pending. For now, the clearest next step for users considering the features is to test them on their own images and tasks, rather than treating Meyer’s limited results or the company’s internal estimates as a production guarantee.
As an affiliate, we earn on qualifying purchases.
Key Questions
What changed in Jev 1.14?
TypeSafe added opt-in image input and a new bounding-box response type. Image input is also supported on Jev 1.13.0, according to TypeSafe.
Is Jev 1.14 ready for production use?
TypeSafe says the new capabilities remain too rough for it to vouch for production use. Image and long-context features require explicit opt-in headers.
What did Meyer’s 30-image test find?
For 10 product images, Meyer reported that Jev 1.14 matched each image to its own headline with scores from 0.65 to 0.89 and identified the product correctly in nine cases. The test was conducted on one website and does not establish broader performance.
When will Jev 1.14 become the default latest version?
TypeSafe said jev-latest was due to move to Jev 1.14 about one week after the October 14 release, around October 21, 2026.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
