Melaya verdict
Melaya is the right pick for operators who need AI agents that go beyond browser tabs and into real Android apps, a capability almost no direct competitor offers at this price point. The platform is still in open beta, and the tool count and subagent depth are impressive on paper, but real-world output quality and reliability at scale remain unverified by independent benchmarks. Teams that only need browser or API-based automation will find the Android-first angle irrelevant and should look at more mature alternatives.
Strengths
Android Device Control operates real phone apps through the accessibility layer, requiring no per-app API, which is a genuine capability gap over most competing agent platforms.
All 7,265+ tools and 111+ subagents are available on every tier, including the free Sandbox, so the tool ceiling does not move as you upgrade.
Weaknesses
Device Control is Android-only with no iOS support listed, which cuts out a large share of mobile-first business users who rely on iPhones.
The platform is in open beta with no published independent benchmarks, no announced funding, and no documented uptime SLA below the Citadel tier, making it a real risk for production-critical workflows.
WEXTL verdict
WEXTL suits small-to-mid-size ops teams that want AI agents woven into multi-step workflows without paying Zapier's per-task rates. The credit model is transparent and the $40/month Grow plan offers real capacity, but the platform only recently left closed beta and has almost no independent user reviews to validate its reliability claims. Teams that need a proven, battle-tested automation backbone should look at n8n or Make before committing here.
Strengths
Credit-based pricing with transparent per-credit costs ($0.000267/credit on Grow) makes execution costs predictable and avoids Zapier-style per-task billing surprises.
Human in the Loop checkpoints are a first-class feature built into the workflow canvas, not a workaround, letting teams insert manual approval gates before AI agents take consequential actions.
Weaknesses
No free tier exists at all, despite the submitted freemium label. A team cannot test a single live workflow without paying at least $19/month, which is a real barrier compared to n8n's self-hosted free option or Make's free tier.
Storage caps are unusually tight: even the $150/month Scale plan allows only 200 MB of total storage and a 1 GB max file size per stream, which will block workflows that process documents, images, or data exports of any real size.
