Pelican-bicycle alternatives
AI
model benchmark
SHOW HN2026
How HN took it
8 positive, 13 neutral or asking, 3 mixed, 10 critical, out of 34 comments about the project. Thread score -0.03.
Execution: 5 for, 7 againstIdea: 1 for, 3 againstDesign: 1 for, 0 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Visual comparison is easier to grasp than numeric scores
From 1 comment
Because it's one of the few ways of comparing models that lets you instantly evaluate them visually. That makes it more comprehensible than a numeric score on a benchmark.
simonw on HN
Some models' SVG outputs are impressive and cute
From 3 comments
Ok the Grok ones are cute
water-drummer on HNI love these.
simonw on HNWhat's interesting to me is that the SVG versions don't have the "AI Image generation hates negative space" issue as badly as generated images do. They kinda stay on point and…
theshrike79 on HN
Objections
Benchmark appears Goodharted or saturated by training on similar SVG tasks
From 3 comments
Website looks very cool, Fable's octopus-organist looks very cute, but I feel like this benchmark (generate an SVG by a short and slightly ridiculous description) in general has…
vova_hn2 on HNWell that benchmark is now saturated, what next. How fast you can hack the pentagon?
neilellis on HNIt’s very likely they all add svg generation into the training data. It’s part of the reason it’s no longer a good benchmark (unless you need to generate SVGs).
gpt5 on HN
Outputs often incoherent, e.g. Gemini Flash results and zebras
From 3 comments
I don't think a bunch of similar tasks can really saturate the "create a SVG of X", because the model should have a quite good spatial understanding of the world and how…
GaggiX on HNBetter than Astra and Fable? It looks quite pretty and even impressive at times if you squint, but look closer and it falls apart in terms of coherency. And I say that as…
Jordan-117 on HNAll models are pretty good now at generating these images. Not zebras. If you want to see how bad SVG output still is, ask for a zebra riding a scooter.
CamperBob2 on HN
Requests
Add Qwen small local models such as Qwen 3.8 Flash-Next
From 2 comments
Would love to see Qwen3.8-27b here, since that is the model most people are running locally.
kennywinker on HNQwen 3.8 Flash-Next is really missing here (because it's small enough to run it locally on a (formerly) affordable machine). This is what Qwen3.8-Flash-Next-UD-Q4_K_XL gives me…
meta-level on HN
Add iterative render-and-fix runs for cheaper models
From 1 comment
I hope there would be parameter iterations allowing especially cheaper models on inspect and fix their output via rendering
miohtama on HN
Questions
Why is this a good test?
From 1 comment
Why is this a good test?
dcreater on HN
Alternatives named
Little Dorrit Benchmark for visual reasoning and structured output
From 1 comment
I'd like to mention the Little Dorrit Benchmark [1] which I have been running for a couple of years now. It has a few nice features: 1. It tests visual reasoning and structured…
svcrunch on HN
MacBook Pro 3D SVG benchmark from playcode.io
From 1 comment
https://playcode.io/blog/macbook-svg-benchmark MacBook Pro 3D in SVG for me the most helpful one.
ianberdin on HN