{ "company": "Tavily", "slug": "tavily", "website": "https://tavily.com", "audit_date": "2026-04-24", "overall_score": 63, "tier": "Human-Dependent", "tier_as_published": "H", "pillars": { "P1": { "name": "Signal Architecture", "score": 18, "max": 25 }, "P2": { "name": "Clarity Stack", "score": 16, "max": 25 }, "P3": { "name": "Trust Envelope", "score": 10, "max": 20 }, "P4": { "name": "Velocity Triggers", "score": 9, "max": 10 }, "P5": { "name": "Gravity Design", "score": 10, "max": 20 } }, "criteria": [ { "id": "P1-A", "pillar": "P1", "name": "Structured Data", "score": 1, "max": 5, "evidence": "No schema.org markup detected. Web search and page inspection returned no JSON-LD, Product, Offer, or AggregateRating schema on the homepage or pricing pages. The site has standard meta tags but nothing structured data agents can parse as authoritative offer signals. Standard SEO metadata only." }, { "id": "P1-B", "pillar": "P1", "name": "Machine-Readable Pricing", "score": 3, "max": 5, "evidence": "Pricing is presented as a clean markdown table inside llms.txt \u2014 plans, credit amounts, and dollar values all machine-readable and in a single retrievable document. However, pricing is not tagged as schema.org/Offer or structured JSON, and the main pricing page appears to be a JavaScript-rendered page not accessible as raw structured data. Solid for LLM retrieval; not yet schema-grade." }, { "id": "P1-C", "pillar": "P1", "name": "llms.txt / Agent Layer", "score": 5, "max": 5, "evidence": "tavily.com/llms.txt is a standout in this audit series. It contains: product descriptions and use-case decision trees (\"when to use Tavily\"), a full pricing table with credits and plan tiers, authentication instructions (Bearer token and OAuth 2.0), all five API endpoints with usage context, CLI install instructions, framework integrations (20+), cloud partnerships (Databricks, AWS, Azure, IBM, Snowflake), and structured external links. This is explicitly written for LLM consumption and is exactly what an agent needs to evaluate, configure, and use the product autonomously." }, { "id": "P1-D", "pillar": "P1", "name": "API / MCP Availability", "score": 5, "max": 5, "evidence": "Full REST API with documented endpoints (Search, Extract, Crawl, Map, Research). Python and JavaScript SDKs (open-source on GitHub). Remote MCP server at mcp.tavily.com supporting OAuth 2.0 for seamless MCP client authentication. Agent Skills CLI for Claude Code, Codex, and Cursor. OpenAPI-spec referenced in docs. This is the most complete agent-integration stack seen in recent audits." }, { "id": "P1-E", "pillar": "P1", "name": "Discoverability (GEO)", "score": 4, "max": 5, "evidence": "robots.txt explicitly names and allows major AI crawlers: ClaudeBot, GPTBot, PerplexityBot, Applebot-Extended, Google-Extended, anthropic-ai, CCBot. llms.txt is at the root domain (high discoverability). Strong developer content presence across LangChain, LlamaIndex, CrewAI, and OpenAI documentation ecosystems. Minor gap: no evidence of structured entity pages or deliberate GEO-specific content strategy beyond llms.txt." }, { "id": "P2-A", "pillar": "P2", "name": "Offer Completeness", "score": 4, "max": 5, "evidence": "All five products (Search, Extract, Crawl, Map, Research), pricing tiers, credit costs, authentication, use cases, and integration options are findable in a single document (llms.txt). Rare for a company to achieve this in one authoritative source. Minor gap: the Research API pricing (async, polling-based) is less clearly defined than the others." }, { "id": "P2-B", "pillar": "P2", "name": "Scope & Limits", "score": 4, "max": 5, "evidence": "Rate limits explicitly stated by environment: 100 RPM (Development, free), 1,000 RPM (Production, paid). Credit costs documented per operation type: 1 credit for basic search, 2 credits for advanced search. Plan credit quantities and monthly caps clearly stated per tier. Enterprise custom limits disclosed as custom. Well above average for API tooling." }, { "id": "P2-C", "pillar": "P2", "name": "Substitution & Fallback Rules", "score": 1, "max": 5, "evidence": "No guidance found on service unavailability, fallback behavior, or what happens when an API call fails or returns empty results. Agents building production workflows cannot pre-validate Tavily's failure behavior without running the API. Standard 0/5 gap for this criteria across most audited companies." }, { "id": "P2-D", "pillar": "P2", "name": "Conditional Logic Transparency", "score": 3, "max": 5, "evidence": "Enterprise plan conditions (SLA, security, custom pricing) require contacting sales \u2014 disclosed in llms.txt, not hidden. Development vs. Production environment distinction (100 RPM vs 1,000 RPM) is documented. Some conditions scattered across docs pages (credit costs by operation, plan upgrade paths). Not all conditions are machine-readable but most are disclosed." }, { "id": "P2-E", "pillar": "P2", "name": "Semantic Precision", "score": 4, "max": 5, "evidence": "Specific, verifiable numbers throughout: \"$0.008/credit,\" \"100 RPM,\" \"1,000 RPM,\" \"1,000 credits/month free,\" \"4,000 credits starting at $0\u2013$99/month.\" Product names (Search, Extract, Crawl, Map, Research) are precisely differentiated with distinct use cases in llms.txt. Minimal vague claims \u2014 strongest at semantic precision of any recent audit." }, { "id": "P3-A", "pillar": "P3", "name": "Verifiable Performance Data", "score": 3, "max": 5, "evidence": "status.tavily.com exists (dedicated status page). trust.tavily.com exists (security and compliance documentation). Insight Partners and Alpha Wave Global $25M Series A provides implicit third-party validation. No G2, Capterra, or Trustpilot presence found (expected for infrastructure/API product). Status page confirms real-time service monitoring, but no public uptime history SLA percentage found." }, { "id": "P3-B", "pillar": "P3", "name": "Scoped Permission Model", "score": 3, "max": 5, "evidence": "OAuth 2.0 for MCP server is a meaningful step toward agent-scoped auth. API key via Bearer token. However, no explicit agent-scoped bounded permissions (time-bounded keys, spending-capped tokens, action-restricted scopes) documented. The OAuth 2.0 support is the strongest signal here \u2014 it enables client-specific auth without sharing master API keys." }, { "id": "P3-C", "pillar": "P3", "name": "Audit Trail / Transaction Log", "score": 2, "max": 5, "evidence": "Credit usage API endpoint is documented (agents can query credit consumption). However, no machine-accessible transaction log, no audit log API, and no webhook system for consumption events found. Usage tracking is human-dashboard oriented based on available documentation." }, { "id": "P3-D", "pillar": "P3", "name": "Behavioral Consistency Signals", "score": 2, "max": 5, "evidence": "Versioned API implied by docs structure. Terms of Service and Privacy Policy exist at tavily.com/terms and tavily.com/privacy. No published changelog, no explicit API version-control policy, no stated notice period for pricing or interface changes found. Platform is young (~2023 founding) \u2014 limited long-term stability track record." }, { "id": "P4-A", "pillar": "P4", "name": "Friction-Free Activation", "score": 5, "max": 5, "evidence": "Free API key at app.tavily.com with no credit card required, 1,000 credits/month automatically. Two-command CLI setup: `pip install tavily-cli && tvly login`. MCP server connection is a single URL with the API key embedded. Agent Skills for Claude Code install automatically via CLI. No human gate at any point. This is textbook frictionless agent activation." }, { "id": "P4-B", "pillar": "P4", "name": "Agent Decision Signals", "score": 4, "max": 5, "evidence": "Free tier provides a clear trial signal for autonomous evaluation. llms.txt explicitly lists \"When to Use Tavily\" \u2014 decision criteria written directly for agent consumption (\"Your agent needs up-to-date information,\" \"RAG pipelines that need grounded sources\"). Minor gap: no explicit upgrade trigger signal (e.g., \"upgrade when credits exhausted this month\" webhook or API endpoint)." }, { "id": "P5-A", "pillar": "P5", "name": "Integration Depth / Switching Cost", "score": 4, "max": 5, "evidence": "Native integrations with LangChain, LlamaIndex, CrewAI, OpenAI Agents SDK, Anthropic Tool Calling, Google ADK, Pydantic AI, Vercel, n8n, Zapier, Flowise, LangFlow, Dify, Composio, and Agno. Enterprise cloud marketplace presence on Databricks, AWS, Azure, IBM watsonx, and Snowflake. Each integration embeds Tavily at the framework level, creating meaningful switching cost \u2014 removing Tavily requires reconfiguring search in every agent framework the team uses." }, { "id": "P5-B", "pillar": "P5", "name": "Agent Memory / Personalization Layer", "score": 2, "max": 5, "evidence": "Credit usage is trackable via API (agents can query their own consumption history). Account-level usage dashboards exist. However, no persistent agent-specific memory layer, no context stored across sessions, no personalized search result tuning based on agent history. Interaction is stateless by design." }, { "id": "P5-C", "pillar": "P5", "name": "Programmatic Renewal Signals", "score": 2, "max": 5, "evidence": "Paid plans auto-renew (standard SaaS billing). No documented API for checking subscription status, querying renewal date, or triggering plan upgrades programmatically. Agents cannot autonomously manage their own Tavily subscription without human intervention." }, { "id": "P5-D", "pillar": "P5", "name": "Compounding Value Signal", "score": 2, "max": 5, "evidence": "The platform's index and retrieval quality improve over time, but no agent-readable signal communicates this. No documented API for accessing cached result quality, search relevance scores over time, or coverage improvement metrics. Value compounds silently \u2014 agents cannot detect or verify it." } ], "strongest_signals": [ { "title": "Best-in-class llms.txt (5/5)", "detail": "Tavily's llms.txt is the most comprehensive agent-identity document seen in this audit series. Pricing, products, use-case decision logic, auth flows, and integrations are all in one machine-readable file \u2014 exactly what a capability-evaluating agent needs before making a procurement decision." }, { "title": "Near-perfect Velocity Triggers (9/10)", "detail": "Free tier with no credit card, two-command CLI install, embedded MCP URL activation, and Agent Skills auto-install create the smoothest onboarding path of any audited company this cycle. An agent system can discover, evaluate, and activate Tavily with zero human intermediation." }, { "title": "Explicit AI crawler allowances", "detail": "robots.txt naming and allowing ClaudeBot, GPTBot, PerplexityBot, and others by name signals a deliberate agent-readiness posture. Most competitors allow by default via wildcard; Tavily explicitly welcomed AI systems." }, { "title": "MCP + OAuth 2.0", "detail": "The combination of a hosted MCP server with OAuth 2.0 support (not just API key auth) is a meaningful step toward secure, client-scoped agent authentication \u2014 rare in this infrastructure category." } ], "critical_gaps": [ { "title": "No schema.org structured data (1/5)", "detail": "The single biggest structural gap. Agents evaluating Tavily via structured data signals \u2014 increasingly the default for agent-native evaluation \u2014 find nothing. The pricing data available in llms.txt is not reflected as parseable schema.org/Offer on the web pages themselves." }, { "title": "Trust Envelope gaps (10/20)", "detail": "No machine-accessible audit log, no scoped agent permissions (bounded by time/amount/action), and no behavioral consistency signals (changelog, versioned terms, notice periods) leave production agent deployments without the trust infrastructure needed for autonomous long-term commitment." }, { "title": "No substitution/fallback rules (1/5)", "detail": "Agents cannot pre-validate what happens when Tavily returns empty results or the service is degraded. Without documented fallback behavior, building reliable agent workflows on Tavily requires defensive engineering on the consumer side." }, { "title": "No compounding value signals (2/5)", "detail": "Agents optimizing their tool stack over time have no mechanism to detect whether Tavily's value has increased \u2014 no retrieval quality metrics, no coverage improvement API, no signal to justify autonomous re-commitment or tier upgrade." } ], "priority_actions": [ { "action": "Add schema.org/Offer markup to pricing page", "points_gain": 3, "pillar": "P1", "effort": "Low" }, { "action": "Add machine-accessible audit/usage log API", "points_gain": 2, "pillar": "P3", "effort": "Medium" }, { "action": "Publish an explicit API changelog + versioning policy", "points_gain": 2, "pillar": "P3", "effort": "Low" }, { "action": "Add agent-scoped API key permissions", "points_gain": 2, "pillar": "P3", "effort": "Medium" }, { "action": "Add substitution/fallback documentation", "points_gain": 2, "pillar": "P2", "effort": "Low" } ], "executive_summary": "Tavily scores 63/100 in the Human-Dependent tier \u2014 technically a classification that undersells its genuine agent-native orientation. The platform's llms.txt is the best-documented agent identity layer in this audit series, its Velocity Triggers score (9/10) is near-perfect, and its robots.txt signals AI-readiness with explicit named allowances for major AI crawlers. Tavily is genuinely built for agents to consume. The tier placement comes down to structural gaps in the trust layer: no schema.org markup on pricing pages, no machine-accessible audit trail, and no agent-scoped permission model mean that autonomous agents cannot fully evaluate, audit, or bound their Tavily usage without human assistance. The single highest-ROI fix \u2014 adding schema.org/Offer to the pricing page \u2014 is a 2-hour engineering task that would push Tavily into the Emerging tier and make its already excellent llms.txt content parseable by schema-aware agent evaluators.", "rubric_version": "v1-2026-04 (20 criteria, 100 raw points; P3-E Agent Registration added to rubric v2 in 2026-06, not scored in this audit)", "framework": "Agent Native Offers \u2014 The Agent Sale framework", "source_file": "2026-04-24 \u2014 Tavily \u2014 Agent Native Offer Audit.md", "rank": 10 }