This product was not featured by Product Hunt yet. It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).
Product upvotes vs the next 3
Waiting for data. Loading
Product comments vs the next 3
Waiting for data. Loading
Product upvote speed vs the next 3
Waiting for data. Loading
Product upvotes and comments
Waiting for data. Loading
Product vs the next 3
Loading
BlazeDiff Agent
Visual regression your coding agent runs
BlazeDiff Agent auto-discovers your routes, captures deterministic screenshots, and runs pixel diffs locally. It hands a small region tile to Claude Code, Cursor, or Codex for a verdict, instead of you eyeballing 40 full-size PNGs. No SaaS. No API key. MIT. Built on N-API (Rust), WASM, and pure-JS diff cores: 3 to 4× faster than odiff, 8× faster than pixelmatch on 4K. Already used by Vitest, Shopify, Ant Design, Vega, and ApexCharts.
Hi everyone, Tim here. Quick story on why I made this.
A few years back, my team had visual tests running in Cypress. They worked, but were slow enough that running them stopped being routine. We moved to a SaaS to fix that: a bit faster, nicer UI, more knobs. But the bill climbed, screenshots left our infra, and we still spent time going through dozens of diffs manually.
I started with the speed part. Wrote a pure-JS pixel-diff faster than pixelmatch, then a Rust+SIMD version via N-API, then a WASM build for browsers. The native binary is now ~3-4x faster than odiff and ~8x pixelmatch on 4K. About half a year in, Vitest, Shopify, and the Ant Group ecosystem ended up pulling BlazeDiff core in.
Next painful part: ambiguous diffs. The tool is local-first, so the question was how to solve this on every engineer's machine. Coding agents have become standard on most teams, so I can just wire it in there.
BlazeDiff Agent uses the same heuristic interpreter I built for CI, crops the diff into small region tiles, and runs a second-pass LLM analysis on just those tiles. It returns a text verdict: regression, intentional, or noise. If the change is a padding shift, it tells you to update the baseline and move on, no clicking. No SaaS, no API key, screenshots never leave the box, MIT.
Two things I'd love feedback on: 1. Does the agent handoff feel like the right primitive, or would you want it fully automated? 2. Anything that would stop you from trying it on your CI today?
BlazeDiff Agent was submitted on Product Hunt and earned 4 upvotes and 1 comments, placing #126 on the daily leaderboard. BlazeDiff Agent auto-discovers your routes, captures deterministic screenshots, and runs pixel diffs locally. It hands a small region tile to Claude Code, Cursor, or Codex for a verdict, instead of you eyeballing 40 full-size PNGs. No SaaS. No API key. MIT. Built on N-API (Rust), WASM, and pure-JS diff cores: 3 to 4× faster than odiff, 8× faster than pixelmatch on 4K. Already used by Vitest, Shopify, Ant Design, Vega, and ApexCharts.
On the analytics side, BlazeDiff Agent competes within Open Source, Developer Tools and GitHub — topics that collectively have 623.8k followers on Product Hunt. The dashboard above tracks how BlazeDiff Agent performed against the three products that launched closest to it on the same day.
Who hunted BlazeDiff Agent?
BlazeDiff Agent was hunted by Teimur Gasanov. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
For a complete overview of BlazeDiff Agent including community comment highlights and product details, visit the product overview.
Teimur Gasanov
Hi everyone, Tim here. Quick story on why I made this.
A few years back, my team had visual tests running in Cypress. They worked, but were slow enough that running them stopped being routine. We moved to a SaaS to fix that: a bit faster, nicer UI, more knobs. But the bill climbed, screenshots left our infra, and we still spent time going through dozens of diffs manually.
I started with the speed part. Wrote a pure-JS pixel-diff faster than pixelmatch, then a Rust+SIMD version via N-API, then a WASM build for browsers. The native binary is now ~3-4x faster than odiff and ~8x pixelmatch on 4K. About half a year in, Vitest, Shopify, and the Ant Group ecosystem ended up pulling BlazeDiff core in.
Next painful part: ambiguous diffs. The tool is local-first, so the question was how to solve this on every engineer's machine. Coding agents have become standard on most teams, so I can just wire it in there.
BlazeDiff Agent uses the same heuristic interpreter I built for CI, crops the diff into small region tiles, and runs a second-pass LLM analysis on just those tiles. It returns a text verdict: regression, intentional, or noise. If the change is a padding shift, it tells you to update the baseline and move on, no clicking. No SaaS, no API key, screenshots never leave the box, MIT.
Two things I'd love feedback on:
1. Does the agent handoff feel like the right primitive, or would you want it fully automated?
2. Anything that would stop you from trying it on your CI today?
Quickstart: `npx @blazediff/agent onboard`.
Repo: https://github.com/teimurjan/blazediff
Around all day, happy to dig into specifics.