The scoreboard
Every AI Agent Readiness Score
One row per product, grouped by letter-grade band and ranked by AI Agent Readiness Score.
The AI Agent Readiness Score rates what an AI agent receives from a product’s public documentation, out of 100 points. Content delivery is worth 50, discovery 25, and agent capabilities 25. Every check is a direct request to the docs; no AI model scores anything. The table groups every product by letter-grade band. A band holds the plus, plain, and minus grades of one letter.
Take the table as data: grades.json (opens in a new tab) grades.csv (opens in a new tab) GitHub repository (opens in a new tab)
Every product by letter-grade band
Filter by category or band, search by name, or sort by any column. Select a column heading again to reverse the order.
Category
Letter-grade band
Showing 97 of 97 products. Grouped by letter-grade band, highest score first.
| Product | Category | Grade | Score | Content delivery points out of 50 | Discovery points out of 25 | Agent capability points out of 25 | Content in server HTML | Markdown mirror | Page fits one agent fetch | llms.txt | MCP server | Platform | Audited |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A band, 90% and higher, 31 products. An agent gets the docs in every format it asks for, finds them through an index, and can use a live MCP server. At most a few points are missing. | |||||||||||||
| Middesk | Identity & auth | A+ | 99.0% | 50/50 | 25/25 | 22/25 | pass | pass | pass | pass | pass | Fern | 2026-10-16 |
| Postman | Cloud & developer platforms | A+ | 99.0% | 50/50 | 25/25 | 22/25 | pass | pass | pass | pass | pass | Fern | 2026-10-16 |
| Clerk | Identity & auth | A+ | 98.0% | 50/50 | 25/25 | 22/25 | pass | pass | pass | pass | pass | Custom Next.js | 2026-10-16 |
| Neon | Databases & analytics | A+ | 97.0% | 50/50 | 25/25 | 21/25 | pass | pass | pass | pass | pass | Custom Next.js | 2026-10-16 |
| Zep | Agent frameworks & memory | A+ | 97.0% | 50/50 | 25/25 | 20/25 | pass | pass | pass | pass | pass | Fern | 2026-10-16 |
| Merge | Payments & business data | A | 96.0% | 50/50 | 25/25 | 19/25 | pass | pass | pass | pass | pass | Fern | 2026-10-16 |
| Persona | Identity & auth | A | 96.0% | 50/50 | 25/25 | 19/25 | pass | pass | pass | pass | pass | Fern | 2026-10-16 |
| Baseten | AI models & inference | A | 95.0% | 50/50 | 25/25 | 19/25 | pass | pass | pass | pass | pass | Mintlify | 2026-10-16 |
| Cartesia | Voice & speech | A | 95.0% | 50/50 | 25/25 | 19/25 | pass | pass | pass | pass | pass | Mintlify | 2026-10-16 |
| Deepgram | Voice & speech | A | 95.0% | 50/50 | 21/25 | 23/25 | pass | pass | pass | fail | pass | Fern | 2026-10-16 |
| Firecrawl | Search & retrieval | A | 95.0% | 50/50 | 25/25 | 19/25 | pass | pass | pass | pass | pass | Mintlify | 2026-10-16 |
| Vercel | Cloud & developer platforms | A | 95.0% | 50/50 | 25/25 | 19/25 | pass | pass | pass | pass | pass | Custom Next.js | 2026-10-16 |
| fal.ai | AI models & inference | A | 94.0% | 50/50 | 25/25 | 18/25 | pass | pass | pass | pass | pass | Mintlify | 2026-10-16 |
| AssemblyAI | Voice & speech | A | 93.0% | 50/50 | 21/25 | 21/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| E2B | Agent sandboxes & browsers | A | 93.0% | 50/50 | 21/25 | 21/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Resend | Email & messaging | A | 93.0% | 50/50 | 17/25 | 25/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Vapi | Voice & speech | A | 93.0% | 50/50 | 21/25 | 20/25 | pass | pass | pass | fail | pass | Fern | 2026-10-16 |
| Metronome | Payments & business data | A- | 92.7% | 50/50 | 21.75/25 | 20/25 | pass | pass | pass | pass | pass | Mintlify | 2026-10-16 |
| Tavily | Search & retrieval | A- | 92.5% | 50/50 | 22.5/25 | 19/25 | pass | pass | pass | pass | pass | Mintlify | 2026-10-16 |
| Reducto | Data pipelines & documents | A- | 92.0% | 50/50 | 21/25 | 20/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| LiveKit | Voice & speech | A- | 91.7% | 50/50 | 23.75/25 | 16/25 | pass | pass | pass | pass | pass | Custom Next.js | 2026-10-16 |
| Blacksmith | Cloud & developer platforms | A- | 91.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Cline | Coding agents | A- | 91.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Doppler | Security & compliance | A- | 91.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | pass | pass | ReadMe | 2026-10-16 |
| Mem0 | Agent frameworks & memory | A- | 91.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Notion | Productivity & collaboration | A- | 91.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Perplexity | Search & retrieval | A- | 91.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Netlify | Cloud & developer platforms | A- | 90.7% | 50/50 | 23.75/25 | 17/25 | pass | pass | pass | pass | pass | Astro Starlight | 2026-10-16 |
| MiniMax | AI models & inference | A- | 90.0% | 50/50 | 21/25 | 18/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Together AI | AI models & inference | A- | 90.0% | 50/50 | 17/25 | 22/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Unstructured | Data pipelines & documents | A- | 90.0% | 50/50 | 21/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| B band, 80% to 89.9%, 22 products. An agent can read and find the docs, but one capability or delivery format is missing or incomplete. | |||||||||||||
| Browserbase | Agent sandboxes & browsers | B+ | 89.7% | 50/50 | 19.75/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| LangChain | Agent frameworks & memory | B+ | 89.7% | 50/50 | 19.75/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Radar | Geospatial & maps | B+ | 89.7% | 50/50 | 19.75/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Exa | Search & retrieval | B+ | 89.0% | 50/50 | 21/25 | 18/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Expo | Cloud & developer platforms | B+ | 89.0% | 50/50 | 21/25 | 17/25 | pass | pass | pass | fail | pass | Custom Next.js | 2026-10-16 |
| SurrealDB | Databases & analytics | B+ | 89.0% | 50/50 | 17/25 | 21/25 | pass | pass | pass | fail | pass | Custom Vike and React | 2026-10-16 |
| Stream | Email & messaging | B+ | 88.5% | 50/50 | 22.5/25 | 15/25 | pass | pass | pass | pass | pass | Custom Next.js | 2026-10-16 |
| Plaid | Payments & business data | B+ | 88.0% | 50/50 | 21/25 | 16/25 | pass | pass | pass | fail | pass | Custom Next.js | 2026-10-16 |
| Runpod | Cloud & developer platforms | B+ | 88.0% | 50/50 | 17/25 | 21/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| WorkOS | Identity & auth | B+ | 88.0% | 45/50 | 21/25 | 21/25 | pass | pass | fail | fail | pass | Custom Next.js | 2026-10-16 |
| Stytch | Identity & auth | B+ | 87.7% | 50/50 | 19.75/25 | 17/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| CodeRabbit | Coding agents | B+ | 87.0% | 50/50 | 17/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| FastMCP | Agent frameworks & memory | B+ | 87.0% | 50/50 | 17/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Amp | Coding agents | B | 86.7% | 50/50 | 19.75/25 | 17/25 | pass | pass | pass | fail | pass | Custom SvelteKit | 2026-10-16 |
| Roboflow | AI models & inference | B | 86.7% | 50/50 | 19.75/25 | 17/25 | pass | pass | pass | fail | pass | GitBook | 2026-10-16 |
| Trigger.dev | Workflows & automation | B | 86.0% | 50/50 | 17/25 | 18/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Honeycomb | Observability & analytics | B | 85.7% | 50/50 | 15.75/25 | 19/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Stripe | Payments & business data | B | 85.7% | 50/50 | 19.75/25 | 16/25 | pass | pass | pass | fail | pass | Custom in-house | 2026-10-16 |
| Courier | Email & messaging | B | 85.0% | 50/50 | 17/25 | 18/25 | pass | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| LlamaIndex | Agent frameworks & memory | B | 84.7% | 50/50 | 19.75/25 | 15/25 | pass | pass | pass | fail | pass | Astro Starlight | 2026-10-16 |
| Tinybird | Databases & analytics | B | 83.0% | 50/50 | 17/25 | 15/25 | pass | pass | pass | fail | pass | Custom Next.js | 2026-10-16 |
| ElevenLabs | Voice & speech | B- | 82.0% | 50/50 | 25/25 | 6/25 | pass | pass | pass | pass | fail | Fern | 2026-10-16 |
| C band, 70% to 79.9%, 24 products. An agent can read the docs, but several checks lose points, often an MCP server, a Markdown version, or an API description. | |||||||||||||
| Supabase | Databases & analytics | C+ | 79.7% | 35/50 | 23.75/25 | 20/25 | pass | fail | pass | pass | pass | Custom Next.js | 2026-10-16 |
| Airbyte | Data pipelines & documents | C+ | 78.0% | 50/50 | 12/25 | 16/25 | pass | pass | pass | fail | pass | Docusaurus | 2026-10-16 |
| Composio | Agent frameworks & memory | C+ | 77.0% | 50/50 | 25/25 | 0/25 | pass | pass | pass | pass | fail | Fumadocs | 2026-10-16 |
| Warp | Coding agents | C+ | 77.0% | 50/50 | 21/25 | 5/25 | pass | pass | pass | fail | fail | Astro Starlight | 2026-10-16 |
| ClickHouse | Databases & analytics | C | 76.7% | 35/50 | 19.75/25 | 21/25 | fail | pass | pass | fail | pass | Mintlify | 2026-10-16 |
| Sentry | Observability & analytics | C | 76.7% | 50/50 | 23.75/25 | 2/25 | pass | pass | pass | pass | fail | Custom Next.js | 2026-10-16 |
| Felt | Geospatial & maps | C | 75.5% | 50/50 | 22.5/25 | 3/25 | pass | pass | pass | pass | fail | GitBook | 2026-10-16 |
| DeepInfra | AI models & inference | C | 75.0% | 50/50 | 21/25 | 3/25 | pass | pass | pass | fail | fail | Mintlify | 2026-10-16 |
| Redis | Databases & analytics | C | 74.7% | 35/50 | 23.75/25 | 16/25 | pass | fail | pass | pass | pass | Hugo | 2026-10-16 |
| Neo4j | Databases & analytics | C | 74.5% | 50/50 | 18.5/25 | 5/25 | pass | pass | pass | fail | fail | Antora and Asciidoctor | 2026-10-16 |
| Recall.ai | Voice & speech | C | 74.0% | 35/50 | 19/25 | 19/25 | pass | fail | pass | fail | pass | ReadMe | 2026-10-16 |
| DuckDB | Databases & analytics | C | 73.7% | 50/50 | 23.75/25 | 0/25 | pass | pass | pass | pass | fail | Jekyll | 2026-10-16 |
| Linear | Productivity & collaboration | C | 73.7% | 35/50 | 23.75/25 | 15/25 | pass | fail | pass | pass | pass | Custom Next.js | 2026-10-16 |
| Railway | Cloud & developer platforms | C | 73.0% | 35/50 | 21/25 | 16/25 | pass | fail | pass | fail | pass | Custom Next.js | 2026-10-16 |
| CARTO | Geospatial & maps | C- | 72.7% | 50/50 | 19.75/25 | 3/25 | pass | pass | pass | fail | fail | GitBook | 2026-10-16 |
| Mailgun | Email & messaging | C- | 72.7% | 35/50 | 20.75/25 | 17/25 | pass | fail | pass | pass | pass | Redocly | 2026-10-16 |
| Massive.com | Payments & business data | C- | 72.7% | 35/50 | 19.75/25 | 17/25 | pass | fail | pass | fail | pass | Custom Next.js | 2026-10-16 |
| n8n | Workflows & automation | C- | 72.7% | 50/50 | 19.75/25 | 3/25 | pass | pass | pass | fail | fail | GitBook | 2026-10-16 |
| Union.ai | Workflows & automation | C- | 72.0% | 50/50 | 21/25 | 1/25 | pass | pass | pass | fail | fail | Hugo | 2026-10-16 |
| CopilotKit | Agent frameworks & memory | C- | 71.7% | 35/50 | 19.75/25 | 16/25 | pass | fail | pass | fail | pass | Custom Next.js | 2026-10-16 |
| Speechmatics | Voice & speech | C- | 71.0% | 50/50 | 19/25 | 0/25 | pass | pass | pass | fail | fail | Docusaurus | 2026-10-16 |
| TrueFoundry | AI models & inference | C- | 71.0% | 50/50 | 17/25 | 3/25 | pass | pass | pass | fail | fail | Mintlify | 2026-10-16 |
| Mapbox | Geospatial & maps | C- | 70.0% | 35/50 | 20/25 | 15/25 | pass | fail | pass | pass | pass | Docusaurus | 2026-10-16 |
| MotherDuck | Databases & analytics | C- | 70.0% | 50/50 | 16/25 | 3/25 | pass | pass | pass | fail | fail | Docusaurus | 2026-10-16 |
| D band, 60% to 69.9%, 6 products. An agent misses major parts of the agent surface, such as both Markdown versions and an MCP server. | |||||||||||||
| Runware | AI models & inference | D+ | 69.7% | 35/50 | 19.75/25 | 15/25 | pass | fail | pass | fail | pass | Custom Astro | 2026-10-16 |
| Mux | Content & web platforms | D+ | 68.5% | 35/50 | 18.5/25 | 15/25 | pass | fail | pass | fail | pass | Custom Next.js | 2026-10-16 |
| Mastra | Agent frameworks & memory | D+ | 68.0% | 50/50 | 16/25 | 2/25 | pass | pass | pass | fail | fail | Docusaurus | 2026-10-16 |
| Inkeep | Agent frameworks & memory | D+ | 67.7% | 50/50 | 16.75/25 | 0/25 | pass | pass | pass | fail | fail | Fumadocs | 2026-10-16 |
| Ory | Identity & auth | D+ | 67.7% | 50/50 | 15.75/25 | 1/25 | pass | pass | pass | fail | fail | Docusaurus | 2026-10-16 |
| Convex | Databases & analytics | D | 64.0% | 35/50 | 13/25 | 15/25 | fail | pass | pass | fail | pass | Docusaurus | 2026-10-16 |
| F band, below 60%, 14 products. An agent receives little it can use without a browser, or most checks lose points. | |||||||||||||
| Letta | Agent frameworks & memory | F | 58.7% | 50/50 | 8.75/25 | 0/25 | pass | pass | pass | fail | fail | Astro Starlight | 2026-10-16 |
| OpenCode | Coding agents | F | 58.7% | 50/50 | 8.75/25 | 0/25 | pass | pass | pass | fail | fail | Astro Starlight | 2026-10-16 |
| PostHog | Observability & analytics | F | 58.5% | 25/50 | 18.5/25 | 15/25 | pass | fail | fail | fail | pass | Gatsby | 2026-10-16 |
| Deno | Cloud & developer platforms | F | 57.0% | 35/50 | 21/25 | 0/25 | pass | fail | pass | fail | fail | Custom Deno Deploy site | 2026-10-16 |
| Hermes Agent | Coding agents | F | 57.0% | 35/50 | 21/25 | 0/25 | pass | fail | pass | fail | fail | Docusaurus | 2026-10-16 |
| Orkes | Workflows & automation | F | 57.0% | 35/50 | 21/25 | 0/25 | pass | fail | pass | fail | fail | MkDocs Material | 2026-10-16 |
| Fused | Geospatial & maps | F | 55.7% | 35/50 | 19.75/25 | 0/25 | pass | fail | pass | fail | fail | Docusaurus | 2026-10-16 |
| Ably | Email & messaging | F | 54.7% | 35/50 | 19.75/25 | 0/25 | pass | fail | pass | fail | fail | Gatsby | 2026-10-16 |
| Daytona | Agent sandboxes & browsers | F | 53.7% | 35/50 | 17.75/25 | 0/25 | pass | fail | pass | fail | fail | Astro Starlight | 2026-10-16 |
| Arcade.dev | Agent frameworks & memory | F | 52.5% | 45/50 | 7.5/25 | 0/25 | pass | pass | fail | fail | fail | Nextra | 2026-10-16 |
| Zilliz Cloud | Search & retrieval | F | 51.2% | 35/50 | 15.25/25 | 0/25 | pass | fail | pass | fail | fail | Docusaurus | 2026-10-16 |
| Overture Maps | Geospatial & maps | F | 49.0% | 35/50 | 13/25 | 0/25 | pass | fail | pass | fail | fail | Docusaurus | 2026-10-16 |
| Vectara | Search & retrieval | F | 35.5% | 20/50 | 15.5/25 | 0/25 | fail | fail | pass | fail | fail | Docusaurus | 2026-10-16 |
| TileDB | Databases & analytics | F | 24.5% | 20/50 | 3.5/25 | 1/25 | fail | fail | pass | fail | fail | Quarto | 2026-10-16 |
No product matches these filters. Choose All or clear the search.
About the columns
- Key checks
- A check mark means the check earns full points. The report shows every check and its evidence.
- Audited
- The date of the audit. The score reflects the documentation on that date.
The latest audit ran on 2026-10-16 with scoring version 100-point-v4. How audit dates work
Products that earn full points, by key check
- Content in server HTML 93 of 97 products
- Markdown mirror 75 of 97 products
- Page fits one agent fetch 94 of 97 products
- llms.txt 28 of 97 products
- MCP server 67 of 97 products
Audit another docs site.
Nominate a docs site