The Agent Readiness Leaderboard
We ranked 500 businesses by their Agent Readiness Score. One scored Gold. Fifty-one scored Silver. Two hundred and fifty scored Bronze. One hundred and ninety-eight are effectively invisible to AI agents. Here is the full picture: who is winning, who is losing, and why the gap between them is wider than most people realize.
The State of Agent Readiness in 2026
The distribution tells the story more clearly than any single number. Out of 500 businesses scanned across every major vertical — SaaS, healthcare, e-commerce, finance, local services, developer tools, communication platforms, and more — the overwhelming majority cluster below the Silver threshold of 60.
The average score is 43 out of 100. The maximum is 75 (Resend). The minimum is 6 (a local restaurant with a PDF menu and a phone number). This 69-point spread between best and worst captures the full range of agent readiness in the current economy: from businesses that AI agents can use autonomously to businesses that might as well not exist in the agent world.
What follows is the data: the top 15 businesses, what they do right, what separates each tier, which industries lead and lag, and the specific patterns that determine where a business lands on this leaderboard.
Top 15 Leaderboard
Only Gold-tier business. Consistent across all 9 dimensions. Lowest dimension: 65 (Payment).
Strong SDKs in 10+ languages. Loses points on pricing complexity and onboarding friction.
Clean developer experience. No agent-card.json or llms.txt. Agent discovery is the gap.
Public status APIs are inherently agent-friendly. Simple, predictable data structures.
Surprisingly strong API docs for a consumer platform. Complex onboarding is the bottleneck.
Has an MCP server (rare). Strong API quality. Agent-native bonus pushes the score.
Excellent API documentation and SDKs. OAuth2 for agent delegation. Pricing complexity hurts.
Best API quality of any business scanned (D2: 90). Pricing transparency (D4: 45) drags the average.
Rich API ecosystem. Onboarding requires human-driven signup. No agent-card.json.
Strong API quality and documentation. Complex pricing calculator hurts D4.
Good API but less focused than Resend. More endpoints, more complexity, lower agent experience.
Strong API and reliability. Wide product surface makes it harder for agents to navigate.
Excellent security dimension. Onboarding complexity and pricing opacity limit the score.
Strong API docs. Usage-based pricing is hard for agents to evaluate without a pricing endpoint.
Clean REST API with good error handling. Limited agent discovery and no MCP server.
Interactive leaderboard with filtering and search at agenthermes.ai/leaderboard. Scores updated continuously as businesses improve and new scans complete.
The 60-Point Cliff: Silver vs. Bronze
The most important number on this leaderboard is not 75 (Resend's Gold score). It is 60 — the threshold between Silver and Bronze. This is where agent utility begins. Below 60, an agent can find your business but cannot use it. Above 60, an agent can discover, evaluate, and begin interacting with your business with reasonable confidence.
The cliff is sharp. The 51 Silver-tier businesses share specific characteristics that the 250 Bronze-tier businesses lack. Understanding what separates them is the key to moving up the leaderboard.
Silver (60-74): What They Have
- Published API documentation with endpoint descriptions
- Structured JSON responses with consistent schemas
- Self-serve signup (even if complex)
- Public pricing page (even if not machine-readable)
- At least one SDK or comprehensive code examples
Bronze (40-59): What They Lack
- API returns HTML errors instead of structured JSON
- No OpenAPI spec or machine-readable API description
- Sales-gated onboarding or manual approval queues
- Pricing hidden behind “Contact Sales”
- No SDKs, no code examples, documentation in prose only
The Bronze Trap
The most dangerous position on the leaderboard is scoring 45-55 — the middle of Bronze. These businesses have enough infrastructure that they appear agent-ready at first glance: they have websites, maybe an API, some documentation. But agents that try to interact with them hit friction at every step. The result is that agents discover these businesses, attempt interaction, fail, and then deprioritize them in future recommendations. A score of 45 is worse than a score of 20 in some ways, because it creates a pattern of failed interactions rather than honest invisibility.
Industry Rankings: Who Leads and Who Lags
Developer Tools / API-first
45 businesses scanned
SaaS / Cloud
85 businesses scanned
Communication Platforms
30 businesses scanned
Finance / Payments
40 businesses scanned
Healthcare
35 businesses scanned
E-commerce
60 businesses scanned
Local Services
90 businesses scanned
Marketing / Advertising
50 businesses scanned
The industry gap is enormous. Developer tools average 58 — nearly Silver. Marketing averages 19 — barely above completely invisible. That 39-point gap between the top and bottom industries is wider than the gap between Silver and Bronze within any single industry.
The pattern is predictable: industries that sell to developers score highest because their product is the API. Industries that sell to consumers or operate in regulated environments (healthcare, finance) score lowest because their infrastructure was built for human interaction and compliance, not for machine-to-machine communication.
But here is the opportunity: the industries with the lowest scores are often the ones where agent-assisted interaction would create the most value. A patient who needs to compare dermatologists, a homeowner who needs to get three plumbing quotes, a small business owner who needs to evaluate marketing agencies — these are high-value agent tasks in low-readiness industries. The first businesses in these verticals to become agent-ready will capture outsized demand.
The Bottom 5: Completely Invisible
At the bottom of the leaderboard, businesses are not just scoring low — they are effectively nonexistent to AI agents. An agent asked to interact with these businesses has no path forward. No API. No structured data. No programmatic interface of any kind. The only option is “call the phone number on the website,” which is not something agents can do.
Website-only presence with phone number. No API, no structured data, no booking system.
Closed ecosystem. No public API for agent interaction. Merchant-facing only.
Consumer app with no developer API. Agents cannot interact at all.
Legacy Square sites have minimal structured data. Modern Square API scores higher when used directly.
PDF menu, phone booking, no pricing online, Facebook page as primary web presence.
Important context: Low scores are not a judgment of business quality. Square has an excellent modern API that scores much higher when accessed directly. Cash App is a great consumer product. Toast runs millions of restaurant transactions. These scores measure agent readiness— how easily AI agents can discover and interact with the business. A score of 8 means “invisible to agents,” not “bad business.”
What Separates Each Tier
After analyzing 500 businesses, clear patterns emerge at each tier boundary. These are not arbitrary thresholds — they map to real capability differences in what agents can accomplish.
Gold (75+)
Consistency across all dimensions. No dimension below 60. An agent can complete the full 6-step journey (find, understand, sign up, connect, use, pay) with high confidence. Only Resend qualifies today.
No weak links in the chain.
Silver (60-74)
Strong in 5-7 dimensions but with 1-2 significant gaps. Agents can interact meaningfully but may need human fallback for onboarding, pricing clarification, or payment. Stripe, GitHub, Vercel, Supabase all live here.
Useful but not autonomous.
Bronze (40-59)
Agents can discover and partially understand the business but cannot complete meaningful tasks. Usually has some API or structured data but critical gaps in onboarding, pricing, or agent experience. The largest tier with 250 businesses.
Visible but not usable.
Not Scored (<40)
Effectively invisible to agents. No public API, no structured data, no programmatic interaction path. The business may have a website, but an agent has no way to extract useful information or take action. 198 businesses — 40% of our dataset — are in this tier.
Invisible. Agents skip these entirely.
Frequently Asked Questions
How often is the leaderboard updated?
We re-scan businesses continuously as our scanner improves and as businesses update their infrastructure. The leaderboard at agenthermes.ai/leaderboard reflects the most recent scan for each business. This blog post captures a snapshot of 500 businesses as of April 2026. Scores can change as businesses improve their agent readiness or as our scoring methodology evolves.
What is the 60-point cliff between Silver and Bronze?
The threshold between Silver (60+) and Bronze (40-59) represents the practical boundary between "an agent can use this business with some friction" and "an agent can find this business but cannot complete a task." Silver-tier businesses have enough structured data, API quality, and documentation that an agent can accomplish basic goals — even if some steps require human fallback. Bronze-tier businesses are visible but functionally unusable by agents. The 60-point line is where agent utility begins.
Why are developer tools and API-first companies at the top?
Developer tools score highest because their product IS an API. They have already solved the problems that other industries struggle with: structured data, clean documentation, programmatic onboarding, and transparent pricing. A business whose primary interface is an API starts with a natural advantage on 5 of our 9 dimensions (D2 API Quality, D6 Data Quality, D7 Security, D8 Reliability, D9 Agent Experience). The remaining dimensions — discovery, onboarding, pricing, and payment — are where even developer tools lose points.
How can my business get on the leaderboard?
Run a free scan at agenthermes.ai/audit. Every business we scan is automatically included in the leaderboard. If you want to improve your position, follow the steps in our improvement guide at agenthermes.ai/blog/improve-agent-readiness-score. The highest-leverage actions are publishing agent-card.json, adding an OpenAPI spec, and making your pricing machine-readable.
Why is Stripe only 68 when it has the best APIs?
Stripe has the highest API Quality score (D2: 90) of any business in our dataset. But the Agent Readiness Score is a weighted average of ALL 9 dimensions, not just API quality. Stripe loses significant points on Pricing Transparency (D4: 45) because their pricing requires parsing complex calculator pages, and on Onboarding (D3) because account setup involves multi-step identity verification. A 45-point spread between best and worst dimensions is what keeps Stripe in Silver instead of Gold. See our full Stripe analysis at agenthermes.ai/blog/why-stripe-scores-68.
Will any business ever reach Platinum (90+)?
Yes, but not yet. Platinum requires capabilities that are still emerging: published A2A agent cards, MCP servers exposing business capabilities, programmatic pricing APIs, OAuth2 for agent delegation, and fully autonomous onboarding. Supabase is closest with their MCP server, and Resend is closest on overall consistency. We expect the first Platinum scores when agent-native protocols gain mainstream adoption — likely late 2026 or 2027.
Where does your business rank?
Get your free Agent Readiness Score and see where you land on the leaderboard. The scan takes 10 seconds and covers all 9 dimensions.