🔍 Read the full analysis: Jev Takes A Look: Testing AI Vision On My Own Images on ThorstenMeyerAI.com
Get the latest gadgets 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, 2026, adding opt-in image support and a bounding-box output type. Thorsten Meyer tested both Jev 1.14 and 1.13 on 30 images from his site; product images matched their headlines well, while the supplied source cuts off before reporting the news-image results and gives no essay-image findings.
Jev is a decision model that responds to typed questions with outputs such as yes-or-no probabilities, selections from a list, or scores. TypeSafe said image input also works with jev-1.13.0, describing that version as “multimodal all along.” The new box output, which returns one bounding box and a confidence score for a target, is available only in Jev 1.14.
Both capabilities require explicit request headers: TypeSafe-Danger-Zone-Images: true for images and an opt-in header for long context. TypeSafe said the safeguards reflect remaining reliability problems, writing that there is “still way too much jaggedness for us to vouch for these for production use cases.” Image inputs accept PNG and JPEG stills, subject to per-image and request-size limits; the release notes say there is no separate image surcharge.
Meyer used the 30 most recent featured images on ThorstenMeyerAI.com: 10 product roundups, 15 news posts and five essays. He asked whether each image fit its own headline, a randomly selected other headline and an unrelated plant-care headline. For the product images, Jev 1.14 scored the matching headline from 0.65 to 0.89, with a median of 0.76; the random headline had a median of 0.06, and the plant headline scored no higher than 0.10. Jev 1.14 also identified nine of 10 products correctly. The source material ends as Meyer begins describing the news-image results, so it does not provide those results or the essay findings.
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
A New Test for Image-Based Decisions
The release gives developers a way to apply Jev’s structured decision format to visual inputs, including asking whether an image suits a headline or locating a named object. That could make the model useful for content review and image classification, where a bounded answer may be easier to evaluate than a free-form description.
Meyer’s product-image results offer an initial example, not a general performance measure. The test used images from one site and a narrow set of prompts; the source does not report a broader benchmark or independent replication. TypeSafe’s own warning that the features are not ready for production use means teams should treat results as experimental and test on their own data before relying on them.
As an affiliate, we earn on qualifying purchases.
How TypeSafe Scoped the Release
Before the October 14 release, Jev’s questions were limited to text, according to Meyer’s account. Jev 1.14 also raises the maximum context limits when users opt in, from 64,000 tokens per request and 32,000 per question to 262,000 for both. TypeSafe still advises against the larger limits because it says reliability declines on long inputs.
TypeSafe described Jev 1.14 as a small release with broad improvements and said multilingual support is expected to benefit most. It made the version available as jev-preview; the company said jev-latest would move to it about a week after release, around October 21. The company also shared internal intelligence estimates rather than public benchmark scores and cautioned that comparisons with open-weight models may be biased in Jev’s favor.
Meyer’s test compared Jev 1.14 with 1.13 using about 60 requests. For product identification, he reported that 1.13 named all 10 products correctly, although two answers had confidence below 0.5. He described the 1.14 error as a plausible mix-up: a mesh Wi-Fi image showed both a router with antennas and a cylindrical mesh unit, and the model selected “Wi-Fi 7 router” with 0.60 confidence.
““was multimodal all along””
— TypeSafe, describing image support in Jev 1.13.0
As an affiliate, we earn on qualifying purchases.
Results Beyond Product Photos
The supplied source cuts off while Meyer is introducing the news-image results. It does not establish how Jev performed on the 15 news images or five essays, whether its outputs were consistent across repeated runs, or how the scores compare with human judgments. The test therefore supports conclusions only about the reported product-image sample.
TypeSafe has not vouched for image support or extended context in production use. The source also does not provide public benchmark scores for the release, and its account of the company’s internal model comparisons says they may favor Jev. Broader accuracy, performance across image types and the practical rate of mistaken bounding boxes remain unconfirmed.
Object detection bounding box software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Version Rollout and Further Testing
TypeSafe said Jev 1.14 was available immediately as jev-preview, with the default jev-latest alias expected to move to it around October 21, 2026. The company’s stated deprecation notice is 60 days for Jev 1.13.0 and 30 days for 1.14.0, with a formal policy still to come.
For now, TypeSafe’s release guidance and Meyer’s limited test point toward further evaluation rather than production reliance. More detail on the news and essay-image results, along with independent tests across different images and tasks, would help show how well the new inputs work beyond this initial site-specific example.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Jev 1.14 add?
TypeSafe added opt-in image input and a box output that returns one bounding box with a confidence score. The box type is limited to Jev 1.14.
Can Jev 1.13 process images?
According to TypeSafe’s release statement, image support also works on jev-1.13.0. Users must opt in using the image request header.
How did Jev perform on the product images?
In Meyer’s test, Jev 1.14 matched product images to their own headlines with scores from 0.65 to 0.89 and identified nine of 10 products correctly. These results come from a small test on one site, not a general benchmark.
Are the image features ready for production?
TypeSafe said it could not vouch for the new capabilities in production use and placed them behind explicit opt-in headers. It advised users to evaluate them for their own use cases.
What is unknown about the test?
The supplied report does not include the results for the 15 news images or five essays, and it gives no independent benchmark or repeated-run assessment.
Source: ThorstenMeyerAI.com
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
![A Walkable ASCII Cyberpunk City In One HTML File [Video]](https://mobquotes.com/wp-content/uploads/2026/09/a-walkable-ascii-cyberpunk-city-in-one-html-file-video-featured-560x300.jpg)
