Meilisearch

# Fix list: Meilisearch

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/meilisearch, the October 2026 research run, assessed 9 October 2026. Grade A, 78.6 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 Meilisearch: 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. Payments & pricing, 40 out of 100, up to 7.5 more on the total

Why it scored 40: Graded on Meilisearch Cloud. No x402, MPP or L402 (0). Unit prices are public. The Build plan includes 50,000 searches and 100,000 documents a month, then $0.40 per 1,000 searches and $0.30 per 1,000 documents, and resource-based projects pay an hourly instance rate plus $0.165 per GiB of disk and $0.15 per GB of bandwidth (20). 14-day trial with no card (20). A person registers in a browser at cloud.meilisearch.com, and no programmatic sign-up was found (0). The self-hosted Community Edition is free under MIT.

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.

## 2. Security & auth, 67 out of 100, up to 5.8 more on the total

Why it scored 67: API keys are limited to named actions and indexes with an optional expiry, tenant tokens embed search rules for each user, keys are sent only in the `Authorization` header and the docs describe rotation (30). A default search-only key, a warning against `*` keys and a read-only MCP route. No confirmation step for destructive calls (15 of 20). Search returns stored documents as written. Guardrail guidance exists for the chat feature, none was found for the MCP route (5 of 15). The Cloud dashboard charts failed and successful API calls, and no audit log was found in the docs (3 of 15). security.txt valid to 26 May 2027, a SECURITY.md that promises a reply within 72 hours, and SOC 2 Type II stated on the pricing page. No bug bounty found, no advisories published on the repository, and the trust centre could not be read (14 of 20).

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.

## 3. Transparency & trust, 74 out of 100, up to 2.3 more on the total

Made of editorial 72, provenance 76.

Why it scored 74: The Community Edition is MIT. Enterprise Edition code, which powers Cloud and adds sharding, is under BUSL-1.1 (27 of 30). The privacy policy of 17 March 2025 names Meili SAS as controller and gives retention periods (the contract term, five years for liability, ten for invoices, three for prospects, six months for navigation data). It sends readers to the trust centre for sub-processors, which we could not read, and no DPA was linked from the pages read. The terms date from 18 March 2022 (20 of 30). A versioning policy commits to semantic versioning and says only the latest release is maintained, with no end-of-life policy. Breaking changes to experimental routes are called out in the changelog (10 of 20). Self-hosted telemetry is documented in full with a `--no-analytics` opt-out, and nine Cloud regions are listed (15 of 20).

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: meilisearch.com, registered 2018-09-19 (8 years) (11 of 15)
- Endpoint on the vendor's domain:  is not on meilisearch.com (0 of 15)
- Terms of service: read, states 4 of the 7 things a reader expects, and has 1 clause that costs points (5.4 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10)

## 4. Agent ergonomics, 87 out of 100, up to 2.1 more on the total

Why it scored 87: `attributesToRetrieve`, `limit`, `cropLength` and `rankingScoreThreshold` size a search response, and the MCP route has four tools, with `describeIndex` capped at five sample documents (22 of 25). `offset` and `limit` or `page` and `hitsPerPage`, filters, facets and sorting (20). Errors return `message`, `code`, `type` and `link` (20). No idempotency keys. Writes by primary key are safe to repeat, and the MCP tool definition in the source has no annotations field, so no readOnlyHint is sent although all four tools only read (10 of 20). Search has no required parameter, and official clients cover JavaScript, Python, PHP, Ruby, Java, Go, .NET and Dart (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.

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

Why it scored 90: Graded as the hosted Meilisearch Cloud service. BetterStack status page with nine regional clusters, the Cloud front end, the Cloud API, the landing page and the docs as components (20). Over 90 days it shows one incident, 16 minutes of downtime on the San Francisco cluster on 24 July 2026 (30). Engine limits are published with numbers (1,000 concurrent searches, 1,000 results a search by default, 10 query words), but no per-plan request rate limit for Cloud was found, although the monitoring docs name rate limiting as a cause of failed calls (8 of 15). A full search queue answers 503 `too_many_search_requests` with `Retry-After` set to 10 seconds, and document writes are keyed by primary key so a repeated write replaces the same document. No idempotency keys (12 of 15). A 99.9 per cent monthly uptime SLA with service credits is published for Enterprise plans (10). The REST API is at 1.54 and generally available. The MCP route is experimental and is not the surface graded here (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, 93 out of 100, up to 1.1 more on the total

Why it scored 93: OpenAPI 3.1.0 file for v1.54.3 with 72 paths and 143 operations, linked from the docs, attached to each release and regenerated in a daily CI check (25). llms.txt on the main site (10). Every operation has a description and search parameters explain defaults and interactions. The four MCP tool descriptions are one line each (16 of 20). 219 schemas, 41 with enums. `filter` is a free-form expression string (12 of 15). Code samples in several languages, 142 of 143 operations document a 4xx response, and an error reference lists 290 codes, each with a type and a docs link (15). Semantic versioning, a dated changelog with an RSS feed and GitHub releases (15).

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, 93 out of 100, up to 0.6 more on the total

Why it scored 93: v1.54.3 tagged 1 October 2026 (30). v1.52 on 3 August, v1.53 on 10 August and v1.54 on 21 September 2026, plus patch releases (20). Of the 14 newest open issues, 11 had at least one reply, and the repository shows 321 open issues and pull requests against 59,529 stars. A Discord server is linked (20 of 25). Official clients are current, with meilisearch-js at 0.62.0 (15). A test suite, SDK tests and a daily OpenAPI check run in GitHub Actions with pinned action versions. We did not confirm the latest run passed (8 of 10).

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.meilisearch.com (sub-processor list, SOC 2 report, DPA). The request for its robots.txt failed at the TLS handshake from our network, so the host was treated as closed
- unchecked: whether Meilisearch is in the official MCP registry
- unchecked: cloud.meilisearch.com sign-up flow and the per-region instance rates, which the docs say appear only in the Cloud console
- unchecked: PyPI download figures and the Python client's latest version (PyPI's robots.txt closes `/pypi/`)
- Whether Cloud enforces a request rate limit, and at what number. The monitoring docs mention one and give no figure
- The pricing page says Cloud starts at $20 a month while its estimator's lowest figures are $23 (resource-based XS) and $30 (Build plan)
- The terms of use forbid comparative analysis of the service. This matters before any probe is run. No deduction taken
- Cloud project hosts are on meilisearch.io, a second domain, which the provenance check may read as off the vendor's domain

## Weaknesses

- The `/mcp` route is an experimental feature that must be switched on, and its four tools carry no read-only annotations
- No per-plan request rate limit is published for Cloud, although the monitoring docs list rate limiting as a cause of failed calls
- The 99.9 per cent uptime SLA applies to Enterprise plans only
- The terms of use date from 18 March 2022 and forbid using the service for comparative analysis, which matters before any probe is run
- Cloud runs the Enterprise Edition, parts of which are under BUSL-1.1, and no audit log was found in the docs

## 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.

- Create a key with only `search`, `documents.get` and `indexes.get` on the indexes needed. The default search key fails `listIndexes` and `describeIndex` over MCP
- Enable the `mcpRoute` experimental feature first. Until then `/mcp` answers `feature_not_enabled`
- For hybrid search send `hybrid` with an `embedder` name configured in the index settings and a `semanticRatio` from 0 to 1
- Writes return a `taskUid` and run asynchronously. Poll the task before searching for the new documents
- On 503 `too_many_search_requests`, wait for the `Retry-After` header (10 seconds) before retrying

## 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.