Read results in context: task, version, harness, date and source ownership.
Anti-bot access success (%)
Reported Mar 21, 2026 · Halluminate BrowserBench run by Browser Use; upstream revision not disclosed
Provider label exactly as reported. Metric measures anti-bot access. COMPUTER USE did not run this evaluation.
Read the original result ↗Methodology →Anti-bot access success (%)
Reported Mar 21, 2026 · Halluminate BrowserBench run by Browser Use; upstream revision not disclosed
Provider label exactly as reported. Metric measures anti-bot access. COMPUTER USE did not run this evaluation.
Read the original result ↗Methodology →Anti-bot access success (%)
Reported Mar 21, 2026 · Halluminate BrowserBench run by Browser Use; upstream revision not disclosed
Provider label exactly as reported. Metric measures anti-bot access. COMPUTER USE did not run this evaluation.
Read the original result ↗Methodology →Anti-bot access success (%)
Reported Mar 21, 2026 · Halluminate BrowserBench run by Browser Use; upstream revision not disclosed
Provider label exactly as reported. Metric measures anti-bot access. COMPUTER USE did not run this evaluation.
Read the original result ↗Methodology →Anti-bot access success (%)
Reported Mar 21, 2026 · Halluminate BrowserBench run by Browser Use; upstream revision not disclosed
Provider label exactly as reported. Metric measures anti-bot access. COMPUTER USE did not run this evaluation.
Read the original result ↗Methodology →Strict task success (%)
$0.17 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Strict task success (%)
$3.40 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Strict task success (%)
$2.20 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Strict task success (%)
$1.55 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Strict task success (%)
$1.10 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Strict task success (%)
$0.62 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Strict task success (%)
$0.44 / solved task (source-reported)
Reported Aug 1, 2026 · Browser Use shared harness, 106 live-web tasks; release/commit not disclosed
Exact model label from source; no precise model API revision disclosed. Historical score is not attached to a current category model profile. Cost in USD per solved task.
Read the original result ↗Methodology →Native application tasks need an environment-specific benchmark and a result that can be checked outside the agent.
Use a compatibility checklist before putting two published scores in the same ranking.
A higher success rate and a lower per-run price describe different parts of the decision.
Identify who designed, executed, scored, and published the evaluation before interpreting the result.
Task success rate (%) · desktop · XLang / OSWorld research team
Task success rate (%) · desktop · XLang / OSWorld research team
Anti-bot access success (%) · browser · Halluminate; run reported by Browser Use
Strict task success (%) · browser · Browser Use