WPLoadTester 7.0 Is Released
WPLoadTester 7.0 is finally released! The beta program that started in February is over, the release build is on the download page, and pricing is live. Thank you to every team that tested against their own applications; you found problems we could not have manufactured, and 7.0 is better for it.
If you missed the beta announcement, the short version: 7.0 is our biggest release since the product debuted in 2000, and it is the release where AI is actually helpful in a load-testing tool, but on your terms, not the industry’s:
- The AI configures your test. Record your application and the AI Assistant works out the auth flow (PKCE, OAuth, form login), correlates the tokens and session state, and proves the fix by replaying end to end. Configuration that took days of staring at HTTP headers now takes minutes. How it works.
- The AI drafts the performance report. A consultant-style bottleneck analysis in which every number traces back to a tool call on your own data. A first draft for your review, not a black box. See the report.
- Load generation runs in your AWS or Azure account. Flat license, compute at cost, no per-VU-hour meter, and test traffic never leaves your cloud boundary. The cost math.
- Bring your own AI key. Anthropic, OpenAI, Azure OpenAI, or AWS Bedrock. Your data routes from your workstation straight to your provider; we are not in the path, and we add no markup. Why BYO-AI matters.
Three things changed since February. First, Azure OpenAI was added to the supported providers, which matters if your company’s AI approval lives in an Azure tenant. Second, the auto-configuration and the Load Test Analytics dashboard were hardened the only way that counts: real teams, real applications, real traffic, including paid consulting engagements run entirely on 7.0. Third, the system now learns.
That last one deserves more of an explanation. When the AI Assistant solves a correlation problem, it writes the fix as a permanent rule in Application State Management, the deterministic expert system that configures recordings before the AI is ever invoked. Solve a problem once, and every future recording gets the fix for free: zero tokens, identical behavior. I wrote up the architecture in The More You Use the AI, the Less It Runs.
The pricing for new customers is live and simpler than the industry’s meters. (All current customers can stick with their current pricing.) The Free edition is $0, forever, on one machine: record, configure with your own AI key, and validate a real test case, with no credit card and no timeout. Cloud load starts at Solo, $2,900 a year ($315 monthly) for up to 5,000 virtual users, then Team at $6,900 for 10,000, Scale at $14,000 for 50,000, and Enterprise beyond that. Every tier is a flat license plus your own cloud compute at cost; a 10,000-user, two-hour run is about $4 of EC2.
Still not in 7.0: real-browser replay. That is 7.1, and until then 7.0 is HTTP protocol only. If a vendor tells you their AI does everything, ask them what is in their next release.
Download the free edition or see the full pricing. If you tried the beta and hit something rough, try the release build. And if you want to talk through whether 7.0 fits your testing problem, get in touch; after 26 years, I still answer our own sales calls.
Founder of Web Performance, Inc
B.S. Electrical & Computer Engineering
The Ohio State University
LinkedIn Profile
Get web performance news
Occasional load-testing tips, performance-engineering notes, and product updates. No spam — unsubscribe anytime.