{ "areas": [ { "blurb": "Model APIs, fast inference and routers. The brain, priced per token.", "categories": [ "inference", "embeddings", "fine-tuning", "gpu-compute", "guardrails", "local-ai", "decision-models" ], "leader": "openai-api", "name": "Models \u0026 inference", "slug": "models", "toolCount": 53, "url": "https://www.anchorterminal.com/tools/?area=models" }, { "blurb": "Models that make images, video and music, from the labs that train them and the platforms that host many. Priced per image, per second or per minute.", "categories": [ "image-generation", "video-generation", "music-generation" ], "leader": "openai-image-api", "name": "Content generation", "slug": "content", "toolCount": 30, "url": "https://www.anchorterminal.com/tools/?area=content" }, { "blurb": "Speech-to-text, text-to-speech, voice agents, phone calling and custom voices. Listed per product, so one company can appear in several places.", "categories": [ "speech-to-text", "text-to-speech", "voice-agents", "voice-calling", "voice-cloning" ], "leader": "twilio-voice", "name": "Voice \u0026 speech", "slug": "voice", "toolCount": 50, "url": "https://www.anchorterminal.com/tools/?area=voice" }, { "blurb": "The loop around the model. Tools, memory, hand-offs, approvals and tracing.", "categories": [ "frameworks", "agent-harnesses" ], "leader": "openai-agents-sdk", "name": "Agent frameworks", "slug": "frameworks", "toolCount": 16, "url": "https://www.anchorterminal.com/tools/?area=frameworks" }, { "blurb": "Where an agent's work runs and what it keeps: sandboxes for code, memory between sessions, the credentials it acts with, and a person to ask.", "categories": [ "code-sandboxes", "agent-memory", "agent-auth", "secrets", "human-in-the-loop" ], "leader": "infisical", "name": "Agent runtime", "slug": "agent-runtime", "toolCount": 37, "url": "https://www.anchorterminal.com/tools/?area=agent-runtime" }, { "blurb": "Search, scraping, data providers, lead and company data, and the APIs that read documents and PDFs.", "categories": [ "search", "web-scrapers", "data-providers", "lead-data", "document-extraction", "pdf-tools" ], "leader": "apify-mcp", "name": "Web, data \u0026 documents", "slug": "web-data", "toolCount": 51, "url": "https://www.anchorterminal.com/tools/?area=web-data" }, { "blurb": "Data and services for particular jobs: travel search and booking, bank data, accounting and weather.", "categories": [ "travel", "banking-data", "accounting", "weather" ], "leader": "apideck-accounting", "name": "Domain data", "slug": "domain-data", "toolCount": 30, "url": "https://www.anchorterminal.com/tools/?area=domain-data" }, { "blurb": "Code hosts, browsers, databases and vector search, observability and agent evals, and cloud, exposed to agents.", "categories": [ "code", "browser", "data", "vector-search", "observability", "agent-observability", "infrastructure", "reasoning" ], "leader": "mongodb-mcp", "name": "Developer \u0026 infrastructure", "slug": "developer", "toolCount": 44, "url": "https://www.anchorterminal.com/tools/?area=developer" }, { "blurb": "Email, agent inboxes, SMS and WhatsApp, and social posting. How an agent reaches people and hears back.", "categories": [ "email", "agent-inboxes", "messaging", "social-media" ], "leader": "twilio", "name": "Communication", "slug": "communication", "toolCount": 30, "url": "https://www.anchorterminal.com/tools/?area=communication" }, { "blurb": "The plain APIs agents reach for all the time: maps and places, translation, calendars, notifications and file storage.", "categories": [ "maps", "translation", "scheduling", "notifications", "file-storage" ], "leader": "google-calendar-api", "name": "Everyday APIs", "slug": "everyday", "toolCount": 36, "url": "https://www.anchorterminal.com/tools/?area=everyday" }, { "blurb": "Design canvases, programmatic image and asset production, and diagramming, exposed to agents.", "categories": [ "design", "design-assets", "diagramming" ], "leader": "figma-mcp", "name": "Design \u0026 diagrams", "slug": "design-diagrams", "toolCount": 18, "url": "https://www.anchorterminal.com/tools/?area=design-diagrams" }, { "blurb": "Work trackers, docs, CRM, support desks, commerce, workflow automation and the services that give agents actions across SaaS apps.", "categories": [ "productivity", "crm", "support", "commerce", "workflow-automation", "aggregator", "company-knowledge" ], "leader": "composio-rube", "name": "Work \u0026 business apps", "slug": "business", "toolCount": 53, "url": "https://www.anchorterminal.com/tools/?area=business" }, { "blurb": "How agents pay and get paid. Pay-per-call and checkout protocols, agent wallets with spending limits, and the platforms that charge agents.", "categories": [ "pay-per-call", "checkout-protocols", "agent-wallets", "payment-platforms" ], "leader": "stripe-mcp", "name": "Payments \u0026 protocols", "slug": "payments", "toolCount": 14, "url": "https://www.anchorterminal.com/tools/?area=payments" } ], "categories": [ { "area": "models", "capabilities": [ "inference.llm", "inference.router", "inference.fast", "inference.open-weights" ], "description": "Frontier and open-weight model APIs, fast inference hardware and routers that sit in front of many providers. Compared on price, context, tool use, data handling and how often models get retired under you.", "indexedCount": 285, "name": "Model APIs \u0026 inference", "slug": "inference", "test": "", "title": "Model APIs and inference for AI agents", "toolCount": 8, "tools": [ "openai-api", "anthropic-api", "groq", "blockrun-ai", "mistral-api", "openrouter", "gemini-api", "deepseek-api" ], "url": "https://www.anchorterminal.com/categories/inference" }, { "area": "models", "capabilities": [ "embed.text", "embed.multimodal", "embed.code", "embed.multilingual", "rerank" ], "description": "Models that turn text, images and code into vectors for search, and rerankers that reorder search results by relevance. Compared on retrieval quality, dimensions and context length, multilingual support and price per million tokens.", "indexedCount": 3, "name": "Embeddings \u0026 rerankers", "slug": "embeddings", "test": "The same corpus and queries embedded with each model, then the top results reranked. We check retrieval quality against labelled answers, latency and the cost per million tokens.", "title": "Embedding and reranking APIs for AI agents", "toolCount": 7, "tools": [ "openai-embeddings", "cohere-embed", "gemini-embedding", "jina-embeddings", "voyage-ai", "mistral-embeddings", "zeroentropy" ], "url": "https://www.anchorterminal.com/categories/embeddings" }, { "area": "models", "capabilities": [ "finetune.sft", "finetune.preference", "finetune.rl", "finetune.lora", "finetune.export" ], "description": "Services that train a model on your examples, by supervised, preference or reinforcement fine-tuning, and serve the result. Compared on which base models you can tune, the methods, price per training token, whether you get the weights and what serving the result costs.", "indexedCount": 0, "name": "Fine-tuning", "slug": "fine-tuning", "test": "The same small dataset used to tune a comparable open model on each service, then served. We check the job flow, how long training takes, whether the weights can leave, and the training and serving cost.", "title": "Fine-tuning services for AI models", "toolCount": 6, "tools": [ "vertex-ai-tuning", "azure-foundry-fine-tuning", "fireworks-fine-tuning", "together-fine-tuning", "unsloth", "tinker" ], "url": "https://www.anchorterminal.com/categories/fine-tuning" }, { "area": "models", "capabilities": [ "compute.gpu", "compute.serverless", "compute.endpoints", "compute.batch", "compute.containers" ], "description": "Clouds that run your own models and jobs on GPUs by the second, as serverless functions, endpoints or rented machines. Compared on GPU types and price per hour, cold starts, scaling and what you have to package.", "indexedCount": 6, "name": "GPU \u0026 serverless compute", "slug": "gpu-compute", "test": "The same model deployed as an endpoint on each platform, called cold and warm, then scaled to zero. We time cold starts, check the scaling and add up the cost per GPU-hour.", "title": "GPU and serverless compute for AI workloads", "toolCount": 9, "tools": [ "modal-sandboxes", "baseten", "modal", "replicate-deploy", "northflank", "beam", "runpod", "lambda", "koyeb" ], "url": "https://www.anchorterminal.com/categories/gpu-compute" }, { "area": "models", "capabilities": [ "guard.injection", "guard.pii", "guard.moderation", "guard.policy", "guard.self-host" ], "description": "APIs and libraries that check what goes into and comes out of a model: prompt injection, jailbreaks, personal data, toxic content and off-policy answers. Compared on what they detect, false positives, the latency they add and where the data goes.", "indexedCount": 8, "name": "Guardrails \u0026 safety filters", "slug": "guardrails", "test": "A set of prompts with injections, jailbreaks, personal data and clean inputs run through each filter. We count what is caught and what is wrongly blocked, and measure the latency each adds.", "title": "Guardrails and safety filters for AI agents", "toolCount": 9, "tools": [ "google-model-armor", "amazon-bedrock-guardrails", "openai-moderation", "nemo-guardrails", "azure-ai-content-safety", "lakera-guard", "mistral-moderation", "guardrails-ai", "galileo" ], "url": "https://www.anchorterminal.com/categories/guardrails" }, { "area": "content", "capabilities": [ "image.generate", "image.edit", "image.upscale", "image.vector" ], "description": "Text-to-image and image editing models behind an API, from the labs that train them and the model platforms that host many of them. Compared on price per image, resolution, edit and reference-image support, text rendering, licence of the output and content rules.", "indexedCount": 6, "name": "Image generation", "slug": "image-generation", "test": "", "title": "Image generation for AI agents", "toolCount": 10, "tools": [ "openai-image-api", "fal-image", "black-forest-labs", "recraft", "ideogram", "replicate-image", "adobe-firefly", "stability-ai-image", "leonardo-ai", "google-imagen" ], "url": "https://www.anchorterminal.com/categories/image-generation" }, { "area": "content", "capabilities": [ "video.generate", "video.image-to-video", "video.edit", "video.audio" ], "description": "Text-to-video and image-to-video models behind an API. Compared on price per second of video, clip length, resolution, audio, how long a job takes to come back and what you're allowed to do with the output.", "indexedCount": 2, "name": "Video generation", "slug": "video-generation", "test": "", "title": "Video generation for AI agents", "toolCount": 10, "tools": [ "alibaba-wan", "runway", "google-veo", "minimax-video", "luma", "pika", "pixverse", "vidu", "kling", "openai-sora" ], "url": "https://www.anchorterminal.com/categories/video-generation" }, { "area": "content", "capabilities": [ "audio.music", "audio.stems", "audio.sfx", "audio.vocals" ], "description": "Models that write music and sound from a prompt, behind an API. Direct providers and model platforms are listed side by side and labelled by kind. Compared on price per minute of audio, track length, stems, vocals, and the licence the output comes with.", "indexedCount": 13, "name": "Music generation", "slug": "music-generation", "test": "", "title": "Music generation for AI agents", "toolCount": 10, "tools": [ "fal-music", "elevenlabs-music", "replicate-musicgen", "stable-audio", "google-lyria", "soundverse", "mubert", "loudly", "beatoven", "soundraw" ], "url": "https://www.anchorterminal.com/categories/music-generation" }, { "area": "voice", "capabilities": [ "speech.stt", "speech.streaming", "speech.batch", "speech.diarisation", "speech.languages" ], "description": "APIs that turn audio into text, streaming or in batch. Compared on accuracy across accents, noise, names and overlapping speakers, streaming latency, languages, diarisation and cost per audio minute.", "indexedCount": 20, "name": "Speech-to-text", "slug": "speech-to-text", "test": "The same audio set through every API, with accents, background noise, proper names and overlapping speakers. We measure word error rate on each slice, streaming latency to a final transcript, and cost per audio minute.", "title": "Speech-to-text APIs for AI agents", "toolCount": 10, "tools": [ "azure-speech-to-text", "amazon-transcribe", "deepgram-stt", "google-speech-to-text", "gladia-stt", "elevenlabs-scribe", "speechmatics-stt", "assemblyai-stt", "soniox-stt", "rev-ai-stt" ], "url": "https://www.anchorterminal.com/categories/speech-to-text" }, { "area": "voice", "capabilities": [ "speech.tts", "speech.streaming", "speech.voices", "speech.ssml", "speech.languages" ], "description": "APIs that turn text into speech, streamed for live conversation or rendered for long-form audio. Compared on time to first audio, pronunciation, how natural it sounds, long-form consistency, languages and cost for a fixed script.", "indexedCount": 4, "name": "Text-to-speech", "slug": "text-to-speech", "test": "One fixed script through every API. We measure time to first audio, check pronunciation of names, numbers and acronyms, judge naturalness blind, listen for drift across a long passage, and price the whole script.", "title": "Text-to-speech APIs for AI agents", "toolCount": 10, "tools": [ "amazon-polly", "azure-text-to-speech", "elevenlabs-tts", "deepgram-tts", "murf-tts", "cartesia-tts", "soniox-tts", "rime-tts", "resemble-ai-tts", "playht-tts" ], "url": "https://www.anchorterminal.com/categories/text-to-speech" }, { "area": "voice", "capabilities": [ "voice.agent", "voice.speech-to-speech", "voice.pipeline", "voice.tools", "voice.telephony" ], "description": "Platforms that run a spoken conversation for you, either with one speech-to-speech model or a configurable pipeline of speech-to-text, a language model and text-to-speech. Compared on response latency, handling interruptions, tool calls, recovery after a misunderstanding and cost per completed call.", "indexedCount": 1, "name": "Conversational voice agents", "slug": "voice-agents", "test": "The same booking and support tasks on every platform. We measure response latency, how interruptions are handled, tool-call success, recovery after a misunderstanding and the total cost per completed call, and record whether the product runs a direct speech-to-speech model or an STT, model and TTS pipeline, since those are different setups.", "title": "Conversational voice-agent APIs", "toolCount": 10, "tools": [ "elevenlabs-agents", "retell-ai", "deepgram-voice-agent", "bland-ai", "vapi", "ultravox", "hume-evi", "bolna", "synthflow", "vogent" ], "url": "https://www.anchorterminal.com/categories/voice-agents" }, { "area": "voice", "capabilities": [ "voice.calls", "voice.numbers", "voice.streaming", "voice.sip", "voice.recording" ], "description": "The calling layer an agent connects to. Numbers, inbound and outbound calls, audio streaming over websockets or SIP, transfers and recordings. Compared on setup, webhook reliability, where numbers are available and the full cost of a call.", "indexedCount": 3, "name": "Phone calling \u0026 voice transport", "slug": "voice-calling", "test": "Inbound and outbound call setup, bidirectional audio streaming, a warm transfer and a recording on every API. We watch webhook reliability, note which countries numbers are available in, and price a five-minute call end to end.", "title": "Phone calling and voice transport APIs for agents", "toolCount": 10, "tools": [ "twilio-voice", "telnyx-voice", "signalwire-voice", "bandwidth-voice", "sinch-voice", "infobip-calls", "plivo-voice", "vonage-voice", "voximplant", "exotel-voice" ], "url": "https://www.anchorterminal.com/categories/voice-calling" }, { "area": "voice", "capabilities": [ "voice.clone", "voice.design", "voice.consent", "speech.tts" ], "description": "APIs that make a new voice from a sample, or design one from a description. Compared on how much audio a clone needs, how close it sounds, consent and verification checks, who owns the voice, and price.", "indexedCount": 0, "name": "Voice cloning \u0026 custom voices", "slug": "voice-cloning", "test": "The same consented sample through every API, short and long. We judge similarity blind, check what consent or verification each vendor asks for, note the licence on the resulting voice, and price a clone and an hour of speech from it.", "title": "Voice cloning and custom voice APIs", "toolCount": 10, "tools": [ "speechify-voice-cloning", "elevenlabs-voice-cloning", "cartesia-voice-cloning", "soniox-voice-cloning", "hume-voice-cloning", "resemble-ai-voice-cloning", "fish-audio-voice-cloning", "murf-voice-cloning", "ultravox-voice-cloning", "playht-voice-cloning" ], "url": "https://www.anchorterminal.com/categories/voice-cloning" }, { "area": "frameworks", "capabilities": [ "agent.framework", "agent.multi-agent", "agent.durable", "agent.mcp-client" ], "description": "Libraries that run the agent loop. Tool calling, MCP, multi-agent hand-offs, durable state, human approval and tracing. Compared on what they do by default, including the telemetry they send.", "indexedCount": 2, "name": "Agent frameworks \u0026 SDKs", "slug": "frameworks", "test": "", "title": "Agent frameworks and SDKs", "toolCount": 6, "tools": [ "openai-agents-sdk", "pydantic-ai", "google-adk", "claude-agent-sdk", "langgraph", "crewai" ], "url": "https://www.anchorterminal.com/categories/frameworks" }, { "area": "agent-runtime", "capabilities": [ "sandbox.code", "sandbox.fs", "sandbox.persist", "sandbox.browser", "sandbox.gpu" ], "description": "Isolated machines an agent can start in seconds to run code, install packages and use a filesystem, then throw away. Compared on start time, isolation, how long a sandbox can live, what it costs per second and whether state can be paused and resumed.", "indexedCount": 10, "name": "Code execution sandboxes", "slug": "code-sandboxes", "test": "The same task in every sandbox: start, install a package, run a script that writes files, pause and resume where supported, then tear down. We time each step, check isolation claims against the docs and add up the cost.", "title": "Code execution sandboxes for AI agents", "toolCount": 7, "tools": [ "modal-sandboxes", "vercel-sandbox", "e2b", "cloudflare-sandbox-sdk", "runloop", "daytona", "blaxel-sandboxes" ], "url": "https://www.anchorterminal.com/categories/code-sandboxes" }, { "area": "agent-runtime", "capabilities": [ "memory.store", "memory.search", "memory.graph", "memory.user", "memory.delete" ], "description": "Services that store what an agent learns about a user or a task and bring back the right parts later: facts, preferences, conversation summaries and knowledge graphs. Compared on recall quality, latency, what they keep and how you delete it.", "indexedCount": 63, "name": "Agent memory", "slug": "agent-memory", "test": "One user's history fed in over several sessions, then questions that need facts from early on, a changed preference and a deleted fact. We check what comes back, how fast, and whether the deletion sticks.", "title": "Memory layers for AI agents", "toolCount": 8, "tools": [ "zep", "honcho", "supermemory", "mem0", "cognee", "memory-reference-server", "graphiti", "hindsight" ], "url": "https://www.anchorterminal.com/categories/agent-memory" }, { "area": "agent-runtime", "capabilities": [ "auth.oauth", "auth.tokens", "auth.consent", "auth.agent-identity", "auth.audit" ], "description": "Services that let an agent act for a user in other apps: OAuth flows and token storage for third-party APIs, scoped and revocable access, and an identity for the agent itself. Compared on the providers they cover, how consent works, token handling and audit logs.", "indexedCount": 23, "name": "Agent auth \u0026 delegated access", "slug": "agent-auth", "test": "An agent connects to two third-party apps for a test user, makes calls, has one scope refused and then the grant revoked. We check the consent flow, where tokens live, what the audit log shows and how revocation reaches the agent.", "title": "Auth and delegated access for AI agents", "toolCount": 13, "tools": [ "descope-agentic-identity", "composio-rube", "scalekit-agentkit", "auth0-ai-agents", "nango", "arcade", "pipedream", "stytch-connected-apps", "workos-pipes", "zapier-mcp", "keycard", "permit-mcp-gateway", "paragon" ], "url": "https://www.anchorterminal.com/categories/agent-auth" }, { "area": "agent-runtime", "capabilities": [ "secrets.store", "secrets.rotate", "secrets.machine-identity", "secrets.audit", "secrets.self-host" ], "description": "Stores for API keys and other secrets that an agent or its runtime reads at call time, instead of keeping them in prompts, config files or environment dumps. Compared on access controls, rotation, audit, SDKs and self-hosting.", "indexedCount": 30, "name": "Secrets \u0026 credential vaults", "slug": "secrets", "test": "An agent runtime reads a key at call time with a scoped machine identity, the key is rotated mid-run and access is then revoked. We check the scoping, how rotation lands, what the audit log records and how long each read takes.", "title": "Secrets managers and credential vaults for agents", "toolCount": 9, "tools": [ "infisical", "aws-secrets-manager", "google-secret-manager", "akeyless", "doppler", "1password", "azure-mcp", "hashicorp-vault", "bitwarden-secrets-manager" ], "url": "https://www.anchorterminal.com/categories/secrets" }, { "area": "agent-runtime", "capabilities": [ "hitl.approve", "hitl.ask", "hitl.handoff", "hitl.channels", "hitl.audit" ], "description": "Services that pause an agent to ask a person: approve an action, answer a question or take over a task, over Slack, email or a web inbox, then resume the agent with the answer. Compared on channels, routing, timeouts and audit.", "indexedCount": 14, "name": "Human approval \u0026 handoff", "slug": "human-in-the-loop", "test": "An agent asks for approval of a risky action over two channels and gets one approval, one rejection and one timeout. We check the routing, what the approver sees, how the answer reaches the agent and what is logged.", "title": "Human approval and handoff for AI agents", "toolCount": 7, "tools": [ "temporal", "trigger-dev", "inngest", "permit-mcp-gateway", "orkes-conductor", "pushary", "gotohuman" ], "url": "https://www.anchorterminal.com/categories/human-in-the-loop" }, { "area": "web-data", "capabilities": [ "web.search", "web.fetch", "search.serp", "search.answer", "search.research" ], "description": "Search engines, SERP APIs and research endpoints that give agents fresh knowledge of the web, plus the MCP servers that wrap them. Compared on price per 1,000 queries, freshness, whether results come back as snippets, full text or a written answer, and what the terms let you keep.", "indexedCount": 12, "name": "Web search APIs", "slug": "search", "test": "", "title": "Web search APIs for AI agents", "toolCount": 11, "tools": [ "tavily-mcp", "you-com-api", "parallel-search-api", "linkup", "brave-search-mcp", "exa-mcp", "valyu", "serpapi", "fetch-reference-server", "searchapi-io", "serper" ], "url": "https://www.anchorterminal.com/categories/search" }, { "area": "web-data", "capabilities": [ "web.scrape", "web.extract", "web.crawl", "scraping.proxies", "scraping.js-render", "scraping.anti-bot", "scraping.actors" ], "description": "Hosted APIs that fetch a page for you, render JavaScript and get past anti-bot checks through rotating proxies, then hand back HTML, Markdown or structured JSON. Also page-to-Markdown readers, crawlers and scraper marketplaces. Compared on price per successful request, success rate on protected sites and what the terms let you do with the output.", "indexedCount": 22, "name": "Web scraping \u0026 crawling", "slug": "web-scrapers", "test": "", "title": "Web scraping and crawling APIs for AI agents", "toolCount": 14, "tools": [ "apify-mcp", "firecrawl-mcp", "zenrows", "spider-cloud", "scrapfly", "scrapegraphai", "olostep", "bright-data", "scrapingbee", "scrapeless", "jina-reader", "scrape-do", "scrapingdog", "scrapingant" ], "url": "https://www.anchorterminal.com/categories/web-scrapers" }, { "area": "web-data", "capabilities": [ "data.markets", "market.crypto", "market.indices", "market.news", "data.company", "data.weather", "data.knowledge" ], "description": "Market, company, weather and reference data behind an API. Scored on the data as well as the API, meaning coverage, freshness, history, method, licence and official standing.", "indexedCount": 67, "name": "Data providers", "slug": "data-providers", "test": "", "title": "Data provider APIs for AI agents", "toolCount": 6, "tools": [ "coingecko-x402-api", "massive", "wikimedia", "open-meteo", "companies-house", "coindesk-data-api" ], "url": "https://www.anchorterminal.com/categories/data-providers" }, { "area": "web-data", "capabilities": [ "lead.search", "lead.enrichment", "email.verification", "email.finder", "data.company" ], "description": "APIs that find prospects, enrich people and companies, and find or verify email addresses. Search, enrichment and email verification are different jobs, so each listing says which it does and the Mode filter separates them.", "indexedCount": 11, "name": "Lead \u0026 company data", "slug": "lead-data", "test": "The same 50 company domains and 50 named prospects through every provider. We measure match rate, field completeness, email validity, stale records, latency and cost per usable record, and score search, enrichment and email verification separately.", "title": "Lead, contact and company data APIs for AI agents", "toolCount": 10, "tools": [ "apollo", "coresignal", "lusha", "leadmagic", "fullenrich", "hunter", "enrich-layer", "crustdata", "dropcontact", "prospeo" ], "url": "https://www.anchorterminal.com/categories/lead-data" }, { "area": "web-data", "capabilities": [ "docs.parse", "docs.ocr", "docs.extract", "docs.tables", "docs.chunk" ], "description": "APIs that turn scans, PDFs, invoices and forms into text, tables and structured fields an agent can use. Compared on text accuracy, table structure, field extraction, page references, processing time and price per page.", "indexedCount": 3, "name": "Document parsing \u0026 extraction", "slug": "document-extraction", "test": "The same scans, invoices, tables and long PDFs through every API. We score text accuracy, table structure, field extraction, page references, processing time and price per page.", "title": "Document parsing, OCR and extraction APIs for AI agents", "toolCount": 10, "tools": [ "reducto", "extend", "mindee", "llamaparse", "mistral-ocr", "adobe-pdf-extract", "veryfi", "unstructured", "nanonets", "pdf-co" ], "url": "https://www.anchorterminal.com/categories/document-extraction" }, { "area": "web-data", "capabilities": [ "pdf.convert", "pdf.merge", "pdf.forms", "pdf.generate", "pdf.extract" ], "description": "APIs that merge, split, convert, fill, sign-ready and generate PDFs. Scored apart from structured extraction, because changing a PDF and reading one are different jobs.", "indexedCount": 39, "name": "PDF tools", "slug": "pdf-tools", "test": "The same set of PDFs through every API. We merge, split, convert to and from Office formats, fill a form and add text, then check the output opens cleanly, keeps its layout, and what each operation costs.", "title": "PDF manipulation and conversion APIs for AI agents", "toolCount": 2, "tools": [ "adobe-pdf-extract", "pdf-co" ], "url": "https://www.anchorterminal.com/categories/pdf-tools" }, { "area": "domain-data", "capabilities": [ "travel.flights", "travel.stays", "travel.booking", "travel.changes", "travel.search" ], "description": "APIs that search and book flights, hotels and other stays, with prices, availability, changes and cancellations. Compared on inventory, whether an agent can complete a booking, how payment is handled and what access needs.", "indexedCount": 36, "name": "Travel \u0026 booking", "slug": "travel", "test": "An agent searches a return flight and a hotel, holds or books where the sandbox allows, then changes and cancels. We check the inventory, the booking and payment steps, the change flow and what getting access took.", "title": "Travel search and booking APIs for AI agents", "toolCount": 7, "tools": [ "duffel", "letsfg", "liteapi", "flightclaw", "expedia-rapid", "booking-demand-api", "hotelbeds" ], "url": "https://www.anchorterminal.com/categories/travel" }, { "area": "domain-data", "capabilities": [ "bank.accounts", "bank.transactions", "bank.identity", "bank.payments", "bank.consent" ], "description": "APIs that connect to a person's bank accounts with their consent to read balances and transactions, check identity and income, and start payments. Compared on country coverage, the consent flow, data freshness and price per connection.", "indexedCount": 10, "name": "Bank data \u0026 open banking", "slug": "banking-data", "test": "A sandbox user connects an account, the agent reads balances and categorised transactions, and the consent is revoked. We check coverage, the consent flow, data quality and what revocation does.", "title": "Open banking and bank data APIs for AI agents", "toolCount": 7, "tools": [ "plaid", "truelayer", "yapily", "enable-banking", "salt-edge", "teller", "gocardless-bank-account-data" ], "url": "https://www.anchorterminal.com/categories/banking-data" }, { "area": "domain-data", "capabilities": [ "accounting.ledger", "accounting.invoices", "accounting.bills", "accounting.reports", "accounting.unified" ], "description": "APIs for bookkeeping and invoicing: ledgers, invoices, bills, expenses and reports, straight from the accounting software or through one API across many. Compared on the objects covered, write access, sandboxes, rate limits and price.", "indexedCount": 32, "name": "Accounting \u0026 invoicing", "slug": "accounting", "test": "An agent creates a customer and an invoice, records a payment, reads a profit and loss report and posts a bill in a sandbox company. We check which steps work, the rate limits and how errors come back.", "title": "Accounting and invoicing APIs for AI agents", "toolCount": 8, "tools": [ "apideck-accounting", "merge-accounting", "xero", "freeagent", "rutter", "invoice-ninja", "quickbooks-online", "freshbooks" ], "url": "https://www.anchorterminal.com/categories/accounting" }, { "area": "domain-data", "capabilities": [ "weather.current", "weather.forecast", "weather.historical", "weather.alerts", "weather.marine" ], "description": "APIs for current weather, forecasts, historical data and alerts, by coordinates. Compared on forecast range, historical depth, how often data updates, the terms of use and price per thousand calls.", "indexedCount": 16, "name": "Weather \u0026 climate data", "slug": "weather", "test": "The same locations queried for current conditions, a 10-day forecast and a year of history. We check what each returns, how often it updates, the licence terms and the cost per thousand calls.", "title": "Weather and climate data APIs for AI agents", "toolCount": 9, "tools": [ "google-weather-api", "pirate-weather", "weatherapi-com", "open-meteo", "visual-crossing", "openweather-one-call", "nws-api", "accuweather-api", "met-office-datahub" ], "url": "https://www.anchorterminal.com/categories/weather" }, { "area": "developer", "capabilities": [ "code.repo", "code.git", "code.docs", "code.deploy", "docs.microsoft" ], "description": "Repository platforms, version control, and up-to-date library documentation delivered straight into the context window.", "indexedCount": 15, "name": "Code \u0026 developer platforms", "slug": "code", "test": "", "title": "Code, repository and documentation tools for AI agents", "toolCount": 5, "tools": [ "context7", "github-mcp-server", "salesforce-dx-mcp", "git-reference-server", "microsoft-learn-mcp" ], "url": "https://www.anchorterminal.com/categories/code" }, { "area": "developer", "capabilities": [ "browser.control", "browser.debug" ], "description": "Servers that let an agent drive a real browser. Navigate, read accessibility trees, click, fill forms, capture traces and screenshots.", "indexedCount": 21, "name": "Browser automation", "slug": "browser", "test": "", "title": "Browser automation tools for AI agents", "toolCount": 4, "tools": [ "chrome-devtools-mcp", "browserbase", "playwright-mcp", "puppeteer-reference-server-archived" ], "url": "https://www.anchorterminal.com/categories/browser" }, { "area": "developer", "capabilities": [ "db.sql", "db.document", "db.admin", "fs.local", "memory.graph" ], "description": "SQL and document databases, managed backends, local file systems and persistent memory.", "indexedCount": 31, "name": "Databases \u0026 files", "slug": "data", "test": "", "title": "Database, file and memory tools for AI agents", "toolCount": 8, "tools": [ "mongodb-mcp", "supabase-mcp", "coinmarketcap-x402-api", "nansen-x402-api", "filesystem-reference-server", "memory-reference-server", "postgres-mcp-pro", "postgres-reference-server-archived" ], "url": "https://www.anchorterminal.com/categories/data" }, { "area": "developer", "capabilities": [ "db.vector", "db.hybrid", "db.fulltext", "db.filters", "db.serverless" ], "description": "Databases an agent can query for retrieval, with vector, keyword and hybrid search. Compared on retrieval accuracy, filters, how fast an update becomes searchable, p95 latency, setup effort and cost at a stated corpus size, with hosted and self-hosted costs reported apart.", "indexedCount": 6, "name": "Retrieval \u0026 vector search", "slug": "vector-search", "test": "The same corpus indexed and the same queries run on every database. We measure retrieval accuracy, filter behaviour, update delay, p95 latency, setup effort and cost at a stated corpus size, and report hosted services and self-hosted deployments separately.", "title": "Vector search and retrieval databases for AI agents", "toolCount": 10, "tools": [ "pinecone", "supabase-mcp", "qdrant", "typesense", "weaviate", "lancedb", "upstash-vector", "milvus-zilliz", "chroma", "epsilla" ], "url": "https://www.anchorterminal.com/categories/vector-search" }, { "area": "developer", "capabilities": [ "observability.errors", "observability.metrics", "observability.logs", "observability.traces", "observability.incidents", "work.oncall" ], "description": "Errors, metrics, logs, traces, monitors and on-call data for agents that debug and run production systems.", "indexedCount": 59, "name": "Observability \u0026 incidents", "slug": "observability", "test": "", "title": "Observability and incident tools for AI agents", "toolCount": 3, "tools": [ "sentry-mcp", "datadog-mcp", "pagerduty-mcp" ], "url": "https://www.anchorterminal.com/categories/observability" }, { "area": "developer", "capabilities": [ "obs.traces", "obs.evals", "obs.prompts", "obs.gateway", "obs.datasets" ], "description": "Tools that trace agent runs, tool calls and model calls, and run evaluations on them. Compared on what appears in a trace, how fast a failure can be found, whether an evaluation can be reproduced, setup time and cost.", "indexedCount": 5, "name": "Agent observability \u0026 evals", "slug": "agent-observability", "test": "The same agent run instrumented in every tool, including a failed tool call and a retry. We check which steps appear in the trace, whether the failure can be found quickly, whether an evaluation can be reproduced, how long setup takes and what it costs.", "title": "Agent tracing, monitoring and evaluation tools", "toolCount": 10, "tools": [ "arize-phoenix", "langfuse", "langsmith", "respan", "braintrust", "laminar", "honeyhive", "galileo", "helicone", "baserun" ], "url": "https://www.anchorterminal.com/categories/agent-observability" }, { "area": "developer", "capabilities": [ "infra.cloudflare", "infra.aws", "infra.azure", "infra.terraform", "infra.cloud" ], "description": "Cloud provider and infrastructure-as-code servers for deploying, inspecting and documenting infrastructure.", "indexedCount": 65, "name": "Cloud \u0026 infrastructure", "slug": "infrastructure", "test": "", "title": "Cloud infrastructure tools for AI agents", "toolCount": 4, "tools": [ "terraform-mcp", "cloudflare-mcp", "azure-mcp", "aws-mcp-servers" ], "url": "https://www.anchorterminal.com/categories/infrastructure" }, { "area": "developer", "capabilities": [ "reasoning.scaffold" ], "description": "Tools that make no external calls and exist to structure an agent's own thinking.", "indexedCount": 0, "name": "Reasoning scaffolds", "slug": "reasoning", "test": "", "title": "Reasoning scaffold tools for AI agents", "toolCount": 1, "tools": [ "sequential-thinking-reference-server" ], "url": "https://www.anchorterminal.com/categories/reasoning" }, { "area": "communication", "capabilities": [ "email.send", "email.inbound", "email.templates", "email.domains", "email.analytics" ], "description": "APIs that send application and transactional email, and often parse inbound mail by webhook. Compared on price per 1,000 emails, deliverability tooling (SPF, DKIM, DMARC, dedicated IPs), inbound handling, sandbox limits and what an agent needs before its first send.", "indexedCount": 64, "name": "Email delivery APIs", "slug": "email", "test": "", "title": "Email delivery APIs for AI agents", "toolCount": 9, "tools": [ "resend", "amazon-ses", "postmark", "mailgun", "sendgrid", "loops", "mailjet", "smtp2go", "brevo" ], "url": "https://www.anchorterminal.com/categories/email" }, { "area": "communication", "capabilities": [ "email.inbox", "email.send", "email.inbound", "email.threads", "email.domains" ], "description": "APIs that give an agent its own email address, so it can send, receive replies and carry a thread on over days. Compared on how an inbox is created, how replies reach the agent (webhooks, websockets, polling), threading, custom domains and price per inbox.", "indexedCount": 0, "name": "Agent inbox APIs", "slug": "agent-inboxes", "test": "", "title": "Email inbox APIs for AI agents", "toolCount": 1, "tools": [ "agentmail" ], "url": "https://www.anchorterminal.com/categories/agent-inboxes" }, { "area": "communication", "capabilities": [ "messaging.sms", "messaging.mms", "messaging.whatsapp", "messaging.rcs", "messaging.verify", "messaging.inbound" ], "description": "APIs for SMS, MMS, WhatsApp, RCS and other customer messaging channels. Email and in-app chat are compared separately. Compared on price per message by country, channel coverage, sender registration (10DLC, WhatsApp business verification), inbound webhooks and what an agent needs before it can send its first message.", "indexedCount": 20, "name": "Messaging APIs", "slug": "messaging", "test": "", "title": "SMS, WhatsApp and messaging APIs for AI agents", "toolCount": 10, "tools": [ "twilio", "bird", "telnyx", "vonage", "bandwidth", "sinch", "plivo", "infobip", "360dialog", "clicksend" ], "url": "https://www.anchorterminal.com/categories/messaging" }, { "area": "communication", "capabilities": [ "social.post", "social.schedule", "social.analytics", "social.comments", "social.media-upload" ], "description": "APIs that publish and schedule posts across social networks from one call, and read back comments and analytics. Compared on which networks they reach, which of those need an app review or a business account, price per connected profile, media support, and whether you can host it yourself.", "indexedCount": 38, "name": "Social media posting APIs", "slug": "social-media", "test": "", "title": "Social media posting and scheduling APIs for AI agents", "toolCount": 10, "tools": [ "late", "buffer", "postiz", "upload-post", "ayrshare", "mixpost", "post-bridge", "publer", "metricool", "oneup" ], "url": "https://www.anchorterminal.com/categories/social-media" }, { "area": "everyday", "capabilities": [ "geo.geocode", "geo.reverse", "geo.places", "geo.routing", "geo.tiles" ], "description": "APIs that turn addresses into coordinates and back, find places, and plan routes with travel times. Compared on coverage, accuracy, what the terms let you store and show, and price per thousand calls.", "indexedCount": 5, "name": "Maps, geocoding \u0026 places", "slug": "maps", "test": "The same addresses in five countries geocoded and reverse geocoded, a place search and a route with traffic. We check accuracy against known coordinates, latency, the storage terms and the cost.", "title": "Maps, geocoding and places APIs for AI agents", "toolCount": 7, "tools": [ "mapbox", "opencage", "google-maps-platform", "stadia-maps", "locationiq", "geoapify", "tomtom" ], "url": "https://www.anchorterminal.com/categories/maps" }, { "area": "everyday", "capabilities": [ "translate.text", "translate.documents", "translate.glossary", "translate.detect", "translate.formality" ], "description": "APIs that translate text and documents between languages, with glossaries, formality control and document formatting kept intact. Compared on quality in the languages you need, glossary support, data retention and price per million characters.", "indexedCount": 7, "name": "Translation", "slug": "translation", "test": "The same texts in six language pairs, one with a glossary and one as a formatted document. We check the glossary terms, the formatting, latency, what the provider keeps and the cost per million characters.", "title": "Machine translation APIs for AI agents", "toolCount": 7, "tools": [ "deepl-api", "azure-translator", "amazon-translate", "google-cloud-translation", "lara-translate", "libretranslate", "lingvanex" ], "url": "https://www.anchorterminal.com/categories/translation" }, { "area": "everyday", "capabilities": [ "calendar.read", "calendar.write", "calendar.availability", "calendar.booking", "calendar.webhooks" ], "description": "APIs that read and write calendars across Google and Microsoft, find free time and book meetings, with booking pages and reminders. Compared on the providers covered, availability logic, webhooks and price per connected account.", "indexedCount": 32, "name": "Calendars \u0026 scheduling", "slug": "scheduling", "test": "An agent finds a free slot across a Google and a Microsoft calendar, books it, moves it and cancels it. We check availability accuracy, time zones, webhook delivery and the cost per connected account.", "title": "Calendar and scheduling APIs for AI agents", "toolCount": 7, "tools": [ "google-calendar-api", "nylas-calendar", "calendly", "microsoft-graph-calendar", "cronofy", "cal-com", "apiroc" ], "url": "https://www.anchorterminal.com/categories/scheduling" }, { "area": "everyday", "capabilities": [ "notify.push", "notify.in-app", "notify.multichannel", "notify.preferences", "notify.digest" ], "description": "APIs that send one message to a person across push, in-app, email, SMS and chat, with preferences, batching and delivery tracking. Compared on channels, workflow logic, user preferences and price per notification.", "indexedCount": 7, "name": "Notifications", "slug": "notifications", "test": "One event sent to a user across in-app, push and email, with a digest and a preference opt-out. We check what arrives where, the batching, the opt-out and the delivery log.", "title": "Notification APIs for AI agents", "toolCount": 7, "tools": [ "novu", "suprsend", "courier", "onesignal", "knock", "ntfy", "pushover" ], "url": "https://www.anchorterminal.com/categories/notifications" }, { "area": "everyday", "capabilities": [ "storage.object", "storage.s3", "storage.share", "storage.drive", "storage.presigned" ], "description": "Object storage and file services an agent can put files in and share them from: S3-compatible buckets, and drives with folders, permissions and share links. Compared on price per gigabyte, egress fees, access controls and S3 compatibility.", "indexedCount": 4, "name": "File storage \u0026 sharing", "slug": "file-storage", "test": "An agent uploads a large file, lists a folder, shares a link that expires and deletes the file. We check the upload path, access controls, link expiry, and the storage and egress costs.", "title": "File storage and sharing APIs for AI agents", "toolCount": 9, "tools": [ "amazon-s3", "google-drive-api", "cloudflare-r2", "supabase-mcp", "backblaze-b2", "box-api", "dropbox-api", "bunny-storage", "tigris" ], "url": "https://www.anchorterminal.com/categories/file-storage" }, { "area": "design-diagrams", "capabilities": [ "design.files", "design.components", "design.canvas", "design.comments", "design.code" ], "description": "Design tools and shared canvases an agent can read from and write to. Files, frames, components, variables, boards and comments. Compared on what the API can change as well as read, design-to-code support, MCP servers and how access is scoped.", "indexedCount": 2, "name": "Design workspaces \u0026 canvases", "slug": "design", "test": "", "title": "Design workspace and canvas APIs for AI agents", "toolCount": 5, "tools": [ "figma-mcp", "miro", "lucid", "framer", "penpot" ], "url": "https://www.anchorterminal.com/categories/design" }, { "area": "design-diagrams", "capabilities": [ "design.templates", "design.render", "image.edit", "design.brand" ], "description": "APIs that fill templates and render images, social graphics, banners, PDFs and video from data, or edit images in bulk. Compared on template tooling, output formats, render speed, price per render and what the licence lets you publish.", "indexedCount": 0, "name": "Programmatic asset production", "slug": "design-assets", "test": "", "title": "Programmatic image and design asset APIs", "toolCount": 5, "tools": [ "canva", "templated", "bannerbear", "adobe-photoshop-api", "placid" ], "url": "https://www.anchorterminal.com/categories/design-assets" }, { "area": "design-diagrams", "capabilities": [ "diagram.create", "diagram.as-code", "diagram.edit", "diagram.export", "diagram.architecture" ], "description": "Tools an agent can use to draw architecture, flow and sequence diagrams, through a REST API, an MCP server, an SDK or a diagram-as-code format. Compared on how the diagram is described, whether it stays editable, export formats and what it costs.", "indexedCount": 6, "name": "Diagramming", "slug": "diagramming", "test": "", "title": "Diagramming APIs and diagram-as-code for AI agents", "toolCount": 10, "tools": [ "miro", "drawio", "tldraw", "lucid", "structurizr", "diagrams-so", "whimsical", "eraser", "cloudviz", "mermaid-chart" ], "url": "https://www.anchorterminal.com/categories/diagramming" }, { "area": "business", "capabilities": [ "work.issues", "work.docs", "work.chat", "work.tickets" ], "description": "Issue trackers, wikis, knowledge bases and chat platforms exposed to agents, almost always behind OAuth and an organisation's existing permissions.", "indexedCount": 21, "name": "Work \u0026 productivity", "slug": "productivity", "test": "", "title": "Issue tracking, docs and chat tools for AI agents", "toolCount": 4, "tools": [ "slack-mcp", "notion-mcp", "atlassian-rovo-mcp", "linear-mcp" ], "url": "https://www.anchorterminal.com/categories/productivity" }, { "area": "business", "capabilities": [ "crm.records", "crm.pipeline", "crm.activities", "crm.search", "crm.webhooks" ], "description": "CRM platforms an agent can work in. Contacts, companies, deals, pipelines, activities and notes, behind an API and often an MCP server. Compared on how much of the record an agent can read and write, search, webhooks, rate limits, how access is scoped and price per seat.", "indexedCount": 10, "name": "CRM \u0026 customer platforms", "slug": "crm", "test": "", "title": "CRM and customer-platform tools for AI agents", "toolCount": 10, "tools": [ "hubspot-mcp", "close", "twenty", "attio", "folk", "salesforce", "pipedrive", "copper", "streak", "freshsales" ], "url": "https://www.anchorterminal.com/categories/crm" }, { "area": "business", "capabilities": [ "support.tickets", "support.conversations", "support.contacts", "support.notes", "support.webhooks" ], "description": "Helpdesks and shared inboxes an agent can work in. Tickets, conversations, contacts, notes and statuses. Compared on permissions, safe handoff to a person, audit trail and API coverage.", "indexedCount": 3, "name": "Customer support \u0026 helpdesk", "slug": "support", "test": "A sandbox queue of ten tickets in every helpdesk. An agent retrieves context, drafts a reply, adds an internal note and changes status. We score permissions, safe handoff, audit trail and API coverage, and never send the test replies to real customers.", "title": "Customer support and helpdesk APIs for AI agents", "toolCount": 10, "tools": [ "intercom", "zendesk", "plain", "front", "help-scout", "chatwoot", "pylon", "gorgias", "freshdesk", "crisp" ], "url": "https://www.anchorterminal.com/categories/support" }, { "area": "business", "capabilities": [ "commerce.products", "commerce.cart", "commerce.checkout", "commerce.orders", "commerce.headless" ], "description": "Platforms with a programmable catalogue, cart, checkout and orders. Compared on API coverage, headless support, webhooks, how an agent can build a cart and check out, hosting model and price.", "indexedCount": 48, "name": "Commerce \u0026 checkout", "slug": "commerce", "test": "The same catalogue and order flow on every platform. An agent searches products, builds a cart, applies a discount and places a test order. We score API coverage, webhooks, error handling, setup effort and platform cost, with no real payments.", "title": "Commerce platforms and checkout APIs for AI agents", "toolCount": 10, "tools": [ "shopify", "woocommerce", "vendure", "saleor", "bigcommerce", "commerce-layer", "medusa", "swell", "elastic-path", "snipcart" ], "url": "https://www.anchorterminal.com/categories/commerce" }, { "area": "business", "capabilities": [ "automation.workflows", "automation.apps", "automation.embedded", "automation.code", "automation.webhooks" ], "description": "Workflow builders an agent can trigger, inspect or build in, from no-code scenarios to code-first scripts and embedded integrations. Compared on connection setup, authorisation, action coverage, retries, logs and cost per completed run.", "indexedCount": 29, "name": "Workflow automation", "slug": "workflow-automation", "test": "One workflow across email, CRM and a database on every platform. We compare connection setup, authorisation, action coverage, retries, logs and the cost per completed run.", "title": "Workflow automation platforms with APIs for agents", "toolCount": 12, "tools": [ "temporal", "trigger-dev", "inngest", "pipedream", "make", "workato", "activepieces", "windmill", "tray", "orkes-conductor", "n8n", "paragon" ], "url": "https://www.anchorterminal.com/categories/workflow-automation" }, { "area": "business", "capabilities": [ "automation.apps", "automation.auth", "automation.actions", "agent.tools" ], "description": "Services that hand an agent ready-made actions across thousands of apps, with the OAuth connections and credentials handled for you. Compared on action coverage, how authorisation works per end user, what the agent sees in its context, logs and price per call. Workflow builders are compared separately.", "indexedCount": 0, "name": "Agent tool access", "slug": "aggregator", "test": "The same task across email, CRM and a database, driven by an agent through each service. We compare connection setup, how each end user authorises, action coverage, retries, logs and the cost per completed run.", "title": "Agent tool access: app actions and authentication for agents", "toolCount": 3, "tools": [ "composio-rube", "zapier-mcp", "letme" ], "url": "https://www.anchorterminal.com/categories/aggregator" }, { "area": "payments", "capabilities": [ "payments.protocol", "payments.x402", "payments.stablecoin", "payments.lightning" ], "description": "Protocols where an API answers with a payment request, the agent pays and retries the call. x402, MPP and L402 compared on who governs them, what rails and assets they settle on, what a payment costs, how long settlement takes and what security researchers have found.", "indexedCount": 0, "name": "Pay-per-call protocols", "slug": "pay-per-call", "test": "", "title": "Pay-per-call payment protocols for AI agents", "toolCount": 3, "tools": [ "mpp", "x402", "l402" ], "url": "https://www.anchorterminal.com/categories/pay-per-call" }, { "area": "payments", "capabilities": [ "payments.protocol", "payments.mandate", "payments.card-token", "payments.checkout" ], "description": "Protocols for an agent buying goods or services on a person's behalf, with mandates that prove what the person approved. AP2 and ACP compared on governance, what a merchant has to build, how consent is captured and how disputes work.", "indexedCount": 0, "name": "Agent checkout protocols", "slug": "checkout-protocols", "test": "", "title": "Agent checkout and mandate protocols", "toolCount": 2, "tools": [ "acp", "ap2" ], "url": "https://www.anchorterminal.com/categories/checkout-protocols" }, { "area": "payments", "capabilities": [ "wallet.onchain", "wallet.custody", "wallet.spend-limits", "payments.x402", "payments.card" ], "description": "Wallets and payment credentials an agent can hold, with limits a person sets. Compared on who holds the keys or funds, per-transaction and daily limits, allow-lists, chains and cards supported, and how an operator revokes access.", "indexedCount": 0, "name": "Agent wallets \u0026 spending controls", "slug": "agent-wallets", "test": "", "title": "Agent wallets and spending controls", "toolCount": 3, "tools": [ "circle-wallets", "coinbase-cdp-agentkit", "privy" ], "url": "https://www.anchorterminal.com/categories/agent-wallets" }, { "area": "payments", "capabilities": [ "payments.card", "payments.x402", "payments.stablecoin", "payments.metering", "payments.checkout", "payments.payouts" ], "description": "Platforms that let builders charge agents, route payments between them or meter usage. Compared on fees, rails (cards, stablecoins, bank transfers), x402 and MPP support, settlement time and what an agent can do without a person in the loop.", "indexedCount": 102, "name": "Payment \u0026 monetisation platforms", "slug": "payment-platforms", "test": "", "title": "Payment and monetisation platforms for agents", "toolCount": 6, "tools": [ "stripe-mcp", "tempo", "nevermined", "crossmint", "payman", "skyfire" ], "url": "https://www.anchorterminal.com/categories/payment-platforms" }, { "area": "models", "capabilities": [ "inference.local", "inference.open-weights", "memory.user", "memory.search", "agent.mcp-client" ], "description": "Software that runs models on hardware the owner keeps, a laptop, a desktop or a home server. Model runners with a local API, chat apps, and personal assistants that work from the owner's own files, mail and records, with no cloud account needed. Compared on what models they run, what hardware they need, what leaves the machine, what an agent can call and the licence.", "indexedCount": 0, "name": "Local AI", "slug": "local-ai", "test": "In this run, public evidence against the published checklist, read as software the owner runs on their own hardware (the local-software lines for Reliability and the self-hosted rule for Payments). When the task suites run, the same small open model, prompts and documents on the same machine through each listing's local API or MCP server. We check the setup steps, tokens per second, memory use, whether answers cite the right file and what a network monitor sees leave the machine.", "title": "Local AI: models and assistants that run on your own hardware", "toolCount": 12, "tools": [ "localai", "screenpipe", "llama-cpp", "lm-studio", "ollama", "anythingllm", "open-webui", "jan", "localghost", "khoj", "gpt4all", "underdog" ], "url": "https://www.anchorterminal.com/categories/local-ai" }, { "area": "business", "capabilities": [ "knowledge.search", "data.catalogue", "data.lineage", "work.docs", "memory.graph" ], "description": "Systems that index what a company knows and holds, its documents, chats and tickets or its tables, pipelines and dashboards, so an agent can find the right one with its owner, its lineage and its permissions attached. Compared on what they connect to, whether they keep each source's access controls, what an agent can query and how they're hosted.", "indexedCount": 0, "name": "Company knowledge \u0026 data catalogues", "slug": "company-knowledge", "test": "A fixed set of questions about one test company's documents and data (who owns a table, where a metric comes from, what a policy says), asked through each listing's agent interface as users with different permissions. We check whether answers cite the right source, whether a user ever sees what they shouldn't, and how long a new document takes to become findable.", "title": "Company knowledge search and data catalogues for AI agents", "toolCount": 8, "tools": [ "glean", "openmetadata", "onyx", "marmot", "atlan", "datahub", "guru", "overclock" ], "url": "https://www.anchorterminal.com/categories/company-knowledge" }, { "area": "models", "capabilities": [ "inference.decision" ], "description": "Models that answer typed questions with probabilities instead of text, a yes or no, one label from a set or a level on a rubric, for routing, triage and checks inside agents and workflows. Compared on accuracy, calibration, context, latency, price and whether the weights are open.", "indexedCount": 0, "name": "Decision models", "slug": "decision-models", "test": "A fixed set of labelled decisions (yes or no, one label from a set, a level on a rubric) sent to every model in the same request shape, with the same states and questions. We check accuracy against the labels, calibration (expected calibration error, Brier score and how often an answer given 0.9 or more is wrong), whether answers move when the option order changes, latency and the cost per 1,000 decisions.", "title": "Decision models for AI agents", "toolCount": 4, "tools": [ "convai-laya", "jaredpalmer-kev", "cloudflare-clef", "typesafe-jev" ], "url": "https://www.anchorterminal.com/categories/decision-models" }, { "area": "frameworks", "capabilities": [ "agent.harness", "agent.mcp-client", "agent.multi-agent" ], "description": "Finished programs that run the agent loop for a person or a pipeline, in a terminal, an editor or the vendor's cloud. They plan, call tools, edit files and run commands, where a framework is a library you build that loop with. Compared on what they ask before acting, what the sandbox and the network allow by default, MCP support, headless use and the telemetry they send.", "indexedCount": 0, "name": "Agent harnesses", "slug": "agent-harnesses", "test": "The same small repository task run headless in each harness with one MCP server attached, first fixing a failing test, then a task that needs the network. We check what it asks before acting, what the sandbox blocks, whether the run stops on its own, what it costs and what leaves the machine.", "title": "Agent harnesses and coding agents", "toolCount": 10, "tools": [ "goose", "openai-codex", "gemini-cli", "openhands", "opencode", "claude-code", "cline", "github-copilot-cli", "aider", "cursor-cli" ], "url": "https://www.anchorterminal.com/categories/agent-harnesses" } ], "count": 63, "meta": { "attribution": "Anchor Terminal (https://www.anchorterminal.com)", "docs": "https://www.anchorterminal.com/docs/", "generatedAt": "2026-10-04", "license": "CC-BY-4.0", "method": "https://www.anchorterminal.com/benchmark/", "methodology": "0.3", "openapi": "https://www.anchorterminal.com/openapi.json", "preview": false, "run": "2026-10-01", "runLabel": "October 2026 research run" } }