Tavily API + MCP by Tavily

# Fix list: Tavily API + MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/tavily-mcp, the October 2026 research run, assessed 1 October 2026. Grade BB, 77.2 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Tavily API + MCP: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Security & auth, 52 out of 100, up to 8.4 more on the total

Why it scored 52: Revocable keys, separate development and production keys, and the Logs API filters by key, so several keys per account work. The hosted MCP adds OAuth but maps it to one dashboard key with no scopes (22). The MCP README and docs lead with `?tavilyApiKey=` in the URL (minus 10). Every tool reads except `tavily_feedback`, which posts scores to Tavily, and there's no read-only toolset (10). The home page says requests pass through 'security, privacy, and content validation layers that block PII leakage, prompt injection, and malicious sources'. That's a claim with no technical detail in the docs, plus a `safe_search` flag for adult content (7). The Logs API (filter by key and endpoint since August 2026) and `include_usage` give per-call visibility (13). SOC 2 and ISO named on the privacy page and home page, and the MCP repo patches vulnerable dependencies about monthly. The trust centre is JavaScript-only for us, there's no security.txt per the 30 September check, and no bug bounty found (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 2. Agent ergonomics, 81 out of 100, up to 3.1 more on the total

Why it scored 81: Six MCP tools in tavily-mcp 0.2.23, but about 18,700 characters of tool definitions in the source, 7,000 of them for the `tavily_feedback` tool. No tool filter found. Raw content, images and answers are opt-in, which keeps API responses small (18). `max_results`, domain include and exclude lists, `time_range`, start and end dates, country, and crawl depth, breadth and limit (20). 432 and 433 tell plan limits from pay-as-you-go limits, and 401, 422 and 429 are documented separately (20). No readOnlyHint or destructiveHint in the MCP source. The tools only read, and failed extracts and maps aren't charged (8). Only `query` or `urls` is required, with official Python and JavaScript SDKs (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 3. Payments & pricing, 75 out of 100, up to 3.1 more on the total

Why it scored 75: Official x402 at x402.tavily.com sells advanced search only, $0.01 in USDC on Base, per the 30 September check. Extract, map, crawl and research aren't sold over x402 (15). Credit costs per endpoint and $0.008 a credit pay as you go, published without a login (20). 1,000 free credits a month, no card (20). Keyless search and extract with the `X-Tavily-Access-Mode: keyless` header, plus the x402 route, so an agent needs no human signup (20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 4. Transparency & trust, 65 out of 100, up to 3.1 more on the total

Made of editorial 47, provenance 82.

Why it scored 65: MCP server MIT and SDKs open, API closed under platform terms (18). The privacy policy (24 November 2025) keeps data for the life of the account with no fixed periods, says query data may be used to improve future responses unless a contract says otherwise, and describes no zero-retention option. No DPA on the privacy page, and the trust centre didn't render for us (15). No deprecation policy or dated deprecation notices found (0). Names Stripe, Google Analytics, PostHog and third-party search index providers such as Google, with US processing under Standard Contractual Clauses (14).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: tavily.com, registered 2023-04-13 (3 years) (7 of 15)
- security.txt: not found (0 of 10)

## 5. Reliability, 90 out of 100, up to 2 more on the total

Why it scored 90: incident.io status page at status.tavily.com with separate API, MCP and website components (20). One incident in the last 90 days, website degraded performance on 17 September 2026, with the API and MCP shown at 100% (30). Published limits, 100 requests a minute on development keys and 1,000 on production, crawl 100 and research 20 on both (15). 429 carries `Retry-After` and the docs say to use that value and the status code rather than parse the message (15). No SLA found in the docs index or terms (0). GA (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 6. Schema & documentation, 89 out of 100, up to 1.8 more on the total

Why it scored 89: Public OpenAPI at docs.tavily.com, and typed JSON Schema on every MCP tool (25). llms.txt with Markdown pages and a 'Start here for agents' section (10). Tool descriptions say what each does and when to reach for it, such as search for facts past the model's cutoff, but none say when not to (14). Enums for `search_depth`, `topic`, `time_range` and `include_domains_mode`, ranges for `max_results` (0 to 20) and `chunks_per_source` (1 to 3), and caps of 300 included and 150 excluded domains. The feedback tool takes free-form `value` fields (13). Error table with examples for 400, 401, 422, 429, 432, 433 and 500 (15). Changelog with monthly entries up to August 2026 and versioned SDKs, but the September SDK parameters (`fetch_timeout`, `cache_fallback`) weren't in it yet, and the API path isn't versioned (12).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 7. Maintenance & community, 85 out of 100, up to 1.3 more on the total

Why it scored 85: tavily-mcp 0.2.23 committed on 2026-09-16, tavily-js 0.7.13 and tavily-python 0.8.4 merged on 2026-09-17 and 18 (30). Several MCP and SDK versions and July and August changelog entries in the last 90 days (20). Maintainers merge pull requests and dependency fixes within days. We didn't read the open issues (15). Listed in the official MCP registry as io.github.tavily-ai/tavily-mcp (15). The MCP repo has no CI workflows and no git tags, though it has a test file and regular security bumps. tavily-python runs tests in CI (5).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: trust.tavily.com content (JavaScript-only), including SOC 2 type, subprocessor list and DPA
- unchecked: open GitHub issues and response times on tavily-mcp
- Whether the hosted MCP at mcp.tavily.com exposes the same six tools as the npm package. Its docs page lists two
- x402 scope and keyless limits rely on the 30 September check

## Weaknesses

- x402 covers advanced search only, not extract, map, crawl or research
- Hosted MCP docs lead with `?tavilyApiKey=` in the URL query string
- No readOnlyHint or destructiveHint on the MCP tools, and the feedback tool's definition is longer than the search tool's
- Privacy policy lets query data improve future responses unless a contract says otherwise, with no zero-retention option found
- No SLA, security.txt or deprecation policy found

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Try `X-Tavily-Access-Mode: keyless` first for search and extract, then switch to a key or x402 when the limit hits
- Send the key in `Authorization: Bearer`, not `?tavilyApiKey=` in the MCP URL
- On 429, wait for `Retry-After`. A 432 or 433 means a spend limit, and retrying won't help
- Run /map before /crawl to bound the crawl, since crawl is billed as map plus extract
- Research is async and runs 4 to 250 credits. Create the task, then poll /research/{request_id}

## What the review panel asked for

- Tool filter for the MCP
- A number for the keyless limit
- git tags on MCP releases
- a written deprecation policy
- x402 on extract and map
- Tool filter for feedback
- List all six tools in docs
- Publish an SLA
- State the keyless limit
- header-only key examples
- zero-retention option
- Keyless MCP
- x402 beyond search
- a tool filter
- mark the source index

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.