Pareto 26.9 vs Mistral Large 4
Pareto 26.9 took 10 min 32 s and $0.34 at list price, Mistral Large 4 took 51 min 22 s and $0.61 at list price. Mistral Large 4 took 4.9× as long and cost 1.8× as much.
Pareto 26.9
Unbiased · $7.50/M out
Mistral Large 4
Mistral · $2.10/M out · 4,680 HN pts
Pelican
Generate an SVG of a pelican riding a bicycle
Watch
Draw a mechanical wristwatch that shows the current local time, with a sweeping seconds hand and a date window. Make it look expensive.
Landing
Build the landing page for Lumen Cycles, a Copenhagen workshop that builds custom cargo bikes to order. Make it look like a real brand you would trust with 4,000 euros.
Breakout
Make a Breakout game I can play right now with the keyboard or the mouse. Give it several levels, a score and some juice.
Logo
Design a logo for Night Heron, an independent bookshop that only opens after dark.
About
Write the opening of the About us page for Patchbay, a three-person studio in Rotterdam that repairs vintage synthesizers. About 150 words.
Patchbay is a small workshop in the heart of Rotterdam, where three people spend their days elbow-deep in the guts of vintage synthesizers. We fix the machines that made the sounds you grew up with—the ones that shaped entire genres and then got left in basements when the tech moved on. But we're not a museum. Our benches are covered in solder, schematics, and the occasional broken key that's been there for two weeks longer than we'd like to admit.
We started this studio because we got tired of seeing classic gear being stripped for parts, or worse, turned into Instagram decor. These instruments deserve better. They were built to make noise, and we want to keep them doing exactly that.
Between the three of us, we've fixed everything from a 1960s Buchla to a beat-up Juno that fell down a flight of stairs. If you've got a synth that's gone quiet, bring it in. We'll bring it back.
Patchbay sits in a former warehouse on Rotterdam's northern edge, where the Maas meets the industrial harbour. Three of us work here—no more, no less—surrounded by oscilloscopes, desoldering stations, and the persistent hum of transformers. We repair vintage synthesizers. That is the whole business.
We are not a museum, and we are not a factory. Every Minimoog, ARP 2600, or Juno that crosses our bench receives the same stubborn attention: traced schematics, replaced capacitors, calibrated voltages, hours spent listening for the note that was never quite right. The machines arrive broken, silent, forgotten. They leave singing.
Rotterdam built its reputation on moving heavy things and keeping them moving. We apply the same pragmatism to circuits that predate us by forty years. If it made sound once, it can make sound again. That is the promise we keep, solder joint by solder joint.
Specs and numbers
| Pareto 26.9 | Mistral Large 4 | |
|---|---|---|
| Maker | Unbiased | Mistral |
| Made in | — | France |
| Released | 13 Oct 2026 | 6 Oct 2026 |
| Price in, per M tokens | $2.50 | $0.70 |
| Price out, per M tokens | $7.50 | $2.10 |
| Context | 262,144 tokens | 1,048,576 tokens |
| Weights | closed | open |
| Hacker News points | — | 4,680 |
| Time for the six prompts | 10 min 32 s | 51 min 22 s |
| List price of the six prompts | $0.34 | $0.61 |
Pareto 26.9 was cheaper: $0.27 less at list price, and less on 3 of the 6 prompts. Pareto 26.9 was faster: 40 min 50 s less in total, and faster on 4 of the 6 prompts.
In the recording, both Breakout games play. Pareto 26.9's landing page was cut off at the output limit (12,000 tokens).
Rebuild this screen experiment: Mistral Large 4's rebuild matched the screenshot at 46%; Pareto 26.9 was not in it. Tea Garden experiment: Mistral Large 4's game passed 2 of 5 checks; Pareto 26.9 was not in it.