Web Agents - Autoppia
SN36AI agents that browse the web and complete tasks autonomously, like a digital assistant that clicks for you
Autonomous web agents that navigate, click, fill forms, and complete workflows on websites they've never seen before. Miners build agents evaluated against the Infinite Web Agents (IWA) benchmark, which dynamically generates web environments so agents can never memorize the test. If your agent can handle anything the web throws at it, it wins.
// Web agents that don't break.
Autoppia is a subnet where miners build autonomous web agents: AI systems that can navigate websites, fill forms, click buttons, and complete complex workflows without human intervention. The IWA benchmark generates dynamic, never-before-seen web environments to test agents, ensuring they genuinely understand how to interact with the web rather than memorizing specific sites.
The simple version: Imagine teaching a robot to use any website: book a flight, fill out a form, compare products, complete a purchase. The catch is that the website changes every time, so the robot can't just memorize the steps. It has to actually understand what it's looking at and make decisions. Autoppia is the competition to build that robot.
Centralized equivalent: Think Anthropic's Computer Use, Adept AI, or Multion, but the agents are built through open competition and tested against infinite, dynamically generated web environments.
How it works:
- Miners build self-contained web agents hosted on GitHub. Agents parse HTML, make intelligent decisions, and produce action sequences to complete assigned tasks. Performance and speed are both essential: agents must complete complex workflows in minimal time.
- Validators generate synthetic web tasks using the IWA benchmark (combining metaprogramming, generative AI, and other techniques). They execute miner action sequences in fresh browser instances, capture snapshots after each action, and run multi-type tests: HTML verification, backend event testing, visual assessment, and LLM-based evaluation.
- The problem it solves: Traditional RPA (Robotic Process Automation) is brittle. Scripts break when websites change. Every interface update requires manual reprogramming.
- The opportunity: The RPA market exceeds $12 billion and is growing fast. Autonomous agents that adapt to interface changes (rather than breaking) are the next evolution.
- The Bittensor advantage: The IWA benchmark generates infinite test environments, preventing overfitting to specific websites. This forces genuinely intelligent web navigation rather than memorized scripts. The open-source pipeline means the best agents are publicly available.
- Traction signals: 1,977 commits across 10 contributors with explosive recent activity: 30-72 commits per week in the last month. Led by Riiveer and Daryxx. Roadmap includes reinforcement learning environments and real-website pipelines. Dev activity score: 6.7/10.
Category: Robotic Process Automation | Centralized Competitor: Anthropic Computer Use, Adept AI, Multion, UiPath, Automation Anywhere
Autoppia's recent development velocity is extraordinary. 30-72 commits per week in the last month makes it one of the most actively developed subnets by commit rate. This burst suggests a major product push.
Mechanism:
The IWA benchmark is Autoppia's key innovation. Dynamic web generation means agents can never game the evaluation by memorizing test sites. Each evaluation creates a fresh web environment, so agents must demonstrate genuine understanding of HTML structure, form semantics, and workflow logic.
The multi-type evaluation (HTML verification, backend events, visual assessment, LLM-based judging) provides comprehensive scoring. An agent that produces the right visual result but breaks the backend, or vice versa, won't score well.
The codebase has 1,977 commits across 10 contributors. Despite the smaller team, the commit rate is among the highest in our coverage. The roadmap includes open-source pipelines (Ridge-style), reinforcement learning environments, and real-website task pipelines.
Market metrics are early-stage. At 14,734 TAO market cap with 1,533 holders, Autoppia is small. Gini of 0.770 indicates concentrated ownership. However, net 7-day inflow of +56 TAO is one of the few positive flows in this batch, suggesting growing interest.
Only 1 active miner is listed, which may reflect the early stage of the network or a winner-takes-all dynamic.
- Price decline: -23% 30-day, -10% 90-day. The market hasn't priced in the development velocity yet.
- Concentrated holdings: Gini of 0.770 is high. A small number of holders dominate.
- 1 active miner: Very early stage for mining competition.
- Centralized competition: Anthropic's Computer Use and Adept AI have enormous funding advantages.
