Benchmark: compare electricity on CHECK24 with an AI agent
On CHECK24, Wield's recipe passed 9 of 9 electricity comparisons with a median of 1.5 seconds and no model cost per run. A browser agent running Claude Sonnet 5.5 passed 5 of 5 runs, with a median of 103 seconds and $0.26 per run. The recipe needs two requests per comparison, and neither side was blocked by the site.
CHECK24 Compare electricity tariffs for a postcode and yearly consumption
Read-only, no loginRuns in your browser only
Parameters (7)
| Metric | Wield recipe | Browser Use + Claude Sonnet 5.5 |
|---|---|---|
| Success | 100% | 100% |
| Median time | 1.5 s | 103 s |
| Cost per run | $0 | $0.263 |
9 runs each, 7 November 2026. How we measure
The task
Both sides ran the same three comparisons. Each one sets a postcode and an annual consumption.
| Query | Postcode | Consumption |
|---|---|---|
| q1 | 10178 | 2,500 kWh |
| q2 | 80331 | 3,500 kWh |
| q3 | 50667 | 1,500 kWh |
Pass check: at least five tariffs, each with a provider, a tariff name and a monthly price.
Results (7 Nov 2026)
| Wield recipe | Browser agent | |
|---|---|---|
| Runs passing the check | 9 of 9 | 5 of 5 |
| Median time per run | 1.5 s | 103.4 s |
| Model cost per run | $0 | $0.26 |
| Requests per run (median) | 2 | 470 (30 to the site) |
| Blocked by the site | 0 runs | 0 of 5 runs |
The browser agent took a median of 13 steps, 95,446 input tokens and 2,860 output tokens per run. Its log records no HTTP 401, 403 or 429 response from the site. Wield had no failed runs. Its fastest run took 1,274 ms and its slowest 6,393 ms.
Why the results differ
The recipe makes two requests per comparison. Each call creates a calculation on CHECK24's server, so the recipe never sets overwrite=true. The result also includes an annual cost without bonuses, because CHECK24 shows prices on mixed bases.
The browser agent fills in the form, waits for results and reads them page by page. Its median of 13 steps means 13 model calls per run. That is where most of its time goes.
Caveats
- Small and preliminary. The benchmark was still running on 7 Nov 2026. Pacing at about 30 site requests per hour keeps counts low on some sides.
- Prices change. The tariffs in these runs are a snapshot from 7 Nov 2026.
- Cost is model cost only. Wield's $0 covers model cost. No model runs in the loop when a recipe runs.
Terms of CHECK24
CHECK24's terms prohibit automated data extraction. The recipe runs one comparison at a time that you start, from your browser, with two requests. Switching providers stays on CHECK24 through the results link. Check the site's terms yourself. Wield's recipes note known restrictions.
FAQ
Does CHECK24 block browser agents?
Not in these runs. The browser agent was blocked by the site in 0 of 5 runs, and its log records no 401, 403 or 429 response. Other postcodes and times may differ, and bot protection can change at any time.
How fast is the Wield recipe for comparing electricity on CHECK24?
The median run took 1.5 seconds (1,544 ms) over 9 runs. The fastest took 1,274 ms and the slowest 6,393 ms. All 9 runs passed the check.
Can I rerun this benchmark?
Yes. The method is on the benchmark page, and the query table above is the full input. The raw per-run logs keep destinations, timing, tokens and checks. They do not keep page contents.