-Content By TechnologyNewswire-
Anchor, the browser infrastructure company for AI agents, today launched Action Index, a shared memory of web workflows that agents have already completed successfully. Instead of reasoning through every click from scratch, agents on Anchor replay a proven path as deterministic code, and the model only steps in where a page has changed.
In Anchor’s evaluation on 300 tasks from the Online-Mind2Web benchmark, a small model with Action Index completed 93.3% of tasks at an average cost of $0.016 per task. A frontier model completed 93% at $1.57 per task. The full 300-task run cost $4.87 with Action Index, compared with $472. Without Action Index, the same small model completed 62.7%.
More from News
- Sanctioned At Home: A New Raw Docuseries, Forensically Dissecting The West’s Most Destructive Policy Weapon
- The Democrats Have Taken Congress. What Does This Mean For America’s Startups And SMEs?
- Say No To AI, Lose Google Search – Is That A Fair Choice For Schools To Face?
- Where Does $6.4 Billion In Stolen Crypto Go And What Does It Take To Track Down The Money?
- Experts Comment: Did The Autumn Budget Give UK Startups And Scale-Ups What They Asked For?
- Microsoft Is Reportedly Offering To Pay Staff To Leave While Pouring Billions Into AI – What Does It Say About Big Tech’s Priorities?
- Tairex Launches Multidisciplinary Human-Agent Collaboration Ecosystem To Advance AI Adoption in Real-World Clinical Settings
- I Woke Up To A Meta Account. When Did I Agree To That?
In a separate paired study on 120 tasks, Action Index reduced 90th-percentile token use by 74%, steps by 75% and wall-clock time by 61%. It also lifted open models: two leading open-weight models rose from 67.7% to 88.7% and from 58.7% to 84.7%.
“Agents today pay to re-solve problems that are already solved,” said Idan Raman, co-founder and CEO of Anchor. “The model isn’t the bottleneck for most web work anymore. Memory is. Action Index means the hundredth run of a workflow costs a fraction of the first, and every successful run makes the next one cheaper.”
-This is a paid press release published via TechnologyNewswire-
