openjev Can we run something like Jev in your browser?

A live, local experiment

Decision model in your browser.

A small Qwen model can either read probabilities for your allowed options without decoding them, or write the same kind of distribution token by token. Pick a size, run both on your own GPU, and measure the difference.

browser onlyno backendyour timings~447 MB model

There is no waitlist! Just try it out ↓

Works best on a laptop or desktop. Phones default to Qwen3 0.6B.

00 / setup

Load the model once

{{ text }}

download / cache—
starts only when you click load
model load—download and prepare
warmup—compile passes for both methods

Weights come from Hugging Face and remain in your browser cache. Inputs never leave this page. First load can take several minutes depending on the network and GPU.

01 / decision

Give it a real choice

Try an example
Allowed options
3 / 20

Both paths receive the same decision. One reads option probabilities directly; the other asks the model to write its option probabilities as JSON text.

your decisionstate + question + options
same local modelQwen3.5 · 0.8B
read logitsA…T probabilities
write tokens{options + probabilities}

02A / direct readout

Choice probabilities

no decoding

Read the model’s choice logits and normalize only across the options you supplied.

waiting for a run
total
—
input
—
output
1 readout

02B / generation

JSON probabilities

token by token

Ask the model to estimate the same displayed-option distribution and write it as JSON. Watch every token arrive.

waiting for a run
first token
—
total
—
input
—
output
—
measured wall-time ratiorun it on your GPU

The methods run sequentially on the same loaded model so they do not contend for one GPU. Direct runs first, then generation.

What these numbers do—and do not—mean

Conditional probabilities. Direct scores are a softmax over only the displayed option tokens. They are not calibrated confidence and do not include every answer the model might prefer.

Small local models. The selectable Qwen sizes keep this browser experiment practical. Their judgment quality is not the Phase 1 4B result and is not claimed to match Jev.

Real local timing. Setup, warmup, prompt preparation, direct execution, first generated token and generation completion are timed with performance.now(). No canned results appear.

Quantized weights. The demo uses MLC's upstream q4f16 Qwen3 build. Quantization can change both quality and speed.