v0.1.29717 Sep 2026·JavaScript & JIT, Memory & reading

Setting up and filling arrays of numbers is now several times faster, and faster than the reference browser

This release speeds up the loops that FILL and initialise arrays of numbers, which run at the start of almost every piece of data or graphics work: laying out an image buffer, generating a sequence, seeding a table of values. Floati already processes several array elements at once on the chip's vector unit for loops that transform one array into another; now it does the same for loops that fill an array from a counter or a simple formula, like setting each pixel or each sample in turn. On these loops the work runs several times sooner, and faster than the reference browser, which still fills them one element at a time. It engages only when Floati can prove every write stays inside the array and that doing a batch at once gives exactly the same result as one at a time, so the data is identical to the last bit. It all runs on your device. This release also quietly fixes an internal inefficiency in how class-based objects were tracked, so code that creates many objects of the same kind uses less memory.

Parity with the reference browser

74overall, as of v0.1.343

Web APIs / DOM surface42

Security architecture66

GPU / graphics acceleration68

DevTools / extensions72

Networking85

Media85

Accessibility85

Real-world site compatibility99

HTML parsing100

CSS parsing + selectors100

Layout: block / flex / grid / tables / writing-modes100

Paint + visual100

JavaScript language100

JavaScript performance (JIT)100

HTML parsing100

What changed 3 changes

JavaScript & JIT	Filling and initialising number arrays is several times faster. The loops that set up an array from a counter or a simple formula (each pixel of an image, each sample of a sound, a generated sequence) now process a batch of elements at once on the processor's vector unit, so this setup work, which runs before almost every data or graphics task, finishes several times sooner and faster than the reference browser, which still does it one element at a time.
JavaScript & JIT	It only speeds up what it can prove is safe. The batched path engages only when Floati has proven every write lands inside the array and that a batch step produces exactly the same result as one-at-a-time, so the array's contents are identical to the last bit; anything it cannot prove runs the ordinary way.
Memory & reading	Class-heavy code uses less memory. An internal fix makes objects created from the same class share one tracked shape instead of a fresh one each time, so a loop that builds many objects of the same kind no longer accumulates bookkeeping, freeing memory and helping later speedups engage more reliably.