update benchmark plot with bu-ultra result (78%)

2026-04-22 17:45:09 +02:00 · 2026-03-24 15:51:44 -07:00
parent c8779d242d
commit 671ca0571c
3 changed files with 1 additions and 1 deletions
--- a/README.md
+++ b/README.md
@@ -107,7 +107,7 @@ We benchmark Browser Use across 100 real-world browser tasks. Full benchmark is
 - You want to self-host and deploy browser agents on your own machines

 **Use [Cloud](https://cloud.browser-use.com) (recommended)**
- Much better agent for complex tasks (see plot above)
+- Our most capable models BU Ultra and BU Max (see plot above)
 - Easiest way to start and scale
 - Best stealth with proxy rotation and captcha solving
 - 1000+ integrations (Gmail, Slack, Notion, and more)
--- a/static/accuracy_by_model_dark.png
+++ b/static/accuracy_by_model_dark.png
--- a/static/accuracy_by_model_light.png
+++ b/static/accuracy_by_model_light.png