Alloy's faster performance reduces the time required to capture products and generate or revise AI prototypes.
Alloy is now 2x faster.
Last year, speed was the biggest bottleneck in AI prototyping. When a capture takes 10 minutes or a big prompt stalls, momentum evaporates. Improving performance was the #1 feedback we heard from you in 2025.
Over the last few weeks, we rebuilt our architecture end-to-end for maximum speed. Capturing pages and iterating is now up to 100% faster.
| Type | Previous average | New average | Improvement |
|---|---|---|---|
| Page capture | 6 min 30 sec | 2 min | 225% faster |
| Simple prompt | 2 min | 35 sec | 243% faster |
| Complex prompt | 6 min | 3 min | 100% faster |
Quickly visualize new ideas, spend less time context switching, and stay in a flow state while exploring new directions.
Less waiting. More doing. An Alloy experience where you feel in control and can focus on creating.
It's available today for all users. We can't wait to hear what you think.
Frequently asked questions
Faster AI prototyping starts with reducing capture, generation, and revision time so teams can test more ideas in the same workday. Alloy improves that full loop, with prototype generation completing twice as fast and large product captures running up to ten times faster.
A speed improvement is most useful when it preserves the context and fidelity of the prototype. Alloy accelerates the same product-aware workflow, so you can keep building from your real interface while getting results sooner.
Performance gains vary by task, so compare the parts of the workflow you use most often. In Alloy's benchmarks, prototype generation is twice as fast, single-page capture is three times faster, and large product captures are up to ten times faster.
