Qdrant API + MCP by Qdrant

# Fix list: Qdrant API + MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/qdrant, the October 2026 research run, assessed 1 October 2026. Grade BB, 75.3 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 Qdrant 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. Payments & pricing, 30 out of 100, up to 8.8 more on the total

Why it scored 30: No x402, MPP or L402 (0). Plans are public, but Standard is "usage-based" with hourly compute, memory and disk charges and no per-unit rate on the page, only a calculator (10). The free cluster needs no card per the 30 September check (20). A person signs up in the browser and creates keys in the console (0).

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. Reliability, 80 out of 100, up to 4 more on the total

Why it scored 80: Better Stack status page at status.qdrant.io with per-region components and history (20). Since 1 July, a network-access configuration incident on 14 August marked partial outage across seven cloud regions with 3 minutes of downtime shown for one, a Cloud UI slowdown of 1 hour 31 minutes on 16 August, a 6-minute Cloud API degradation on 21 September and two maintenance windows. Nothing we can read as an hour of core API down, so minor only (20). No published request limits for Qdrant Cloud. Strict mode lets the operator set read and write rate limits per collection, so a mechanism with no vendor numbers (5 of 15). Rate-limited requests return 429 with `Retry-After` in seconds (from the server source), and `wait=true` plus upserts by point ID make writes safe to repeat (15). SLA of 99.5% on Free and Standard, 99.9% to 99.95% with high availability, 99.9% on Premium (10). 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.

## 3. Security & auth, 83 out of 100, up to 3 more on the total

Why it scored 83: Qdrant Cloud database keys can be read-only or read-write, limited to chosen collections, and expire after 90 days by default, and they travel in the `api-key` header or as Bearer (30). Read-only keys and the MCP server's read-only mode, but no confirmation step for deletes (15). Stored payloads come back as written and we found no prompt-injection guidance for the MCP memory tools (5). Audit logging on paid clusters records the operation, user or key, time, collection and result (15). SOC 2 Type 2 and HIPAA per the security page, a Drata trust centre and a bug bounty page. No SECURITY.md in the repository and no security.txt per the 30 September check (18).

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.

## 4. Schema & documentation, 87 out of 100, up to 2.1 more on the total

Why it scored 87: OpenAPI file in the repository at docs/redoc/master/openapi.json, last changed 26 August 2026 (25). llms.txt indexes 547 pages served as Markdown (10). API reference descriptions say what each endpoint does. The MCP find tool says when to use it, the store tool only says "when you are asked to remember something", and neither says when not to (12). OpenAPI types with enums and required fields, but the MCP store tool takes metadata as "any json" (12). Examples on most docs pages and a common-errors page (13). Semver releases with notes on GitHub (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.

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

Why it scored 87: The MCP server has 2 tools, `qdrant-find` and `qdrant-store` (25). `limit`, `offset`, scroll pagination, `with_payload` and `with_vector` to trim responses, and payload filters (20). Errors come back with an HTTP status and a JSON message, mapped per error type, and a common-errors page explains the usual ones (15). Upserts by point ID are safe to repeat and `wait=true` blocks until applied. `QDRANT_READ_ONLY=true` drops the store tool, but neither tool sets readOnlyHint or destructiveHint (12). Few required fields and official clients for Python, TypeScript, Rust, Go, Java and .NET (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.

## 6. Transparency & trust, 84 out of 100, up to 1.4 more on the total

Made of editorial 82, provenance 86.

Why it scored 84: Apache-2.0 (30). The privacy policy names Qdrant Solutions GmbH in Berlin, gives a 90-day limit on IP logs, names processors and transfer bases, but links no DPA or subprocessor page (22). An upgrade policy (one minor version at a time, clients compatible with the last three minors) but no deprecation notice period (10). Self-hosted Qdrant sends anonymised usage statistics by default, documented with an opt-out (`telemetry_disabled` or `--disable-telemetry`), and the security page says cluster data stays in its deployment region (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: qdrant.tech, registered 2020-10-27 (5 years) (11 of 15)
- security.txt: not found (0 of 10)

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

Why it scored 85: Server v1.19.1 tagged 3 September 2026 and the Python client 1.19.1 on 16 September, both within 30 days (30). v1.18.3, v1.19.0 and v1.19.1 since 3 July (20). The dev branch took a fix on 1 October, but 474 issues are open and a batch of bug reports from 24 July still sits open, and we couldn't see reply counts (15). Official clients in six languages, current with the server (15). 18 CI workflows including lint, tests, integration and coverage, plus Dependabot (10). Less 5 because the MCP server's last release was v0.8.1 on 10 December 2025.

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.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- GHSA-f632-vm87-2m2f, high severity, arbitrary file write through the `/logger` endpoint. The fix ("Restrict /logger API", #7527) landed on 13 November 2025 and shipped in v1.16.0, and the advisory was published on 5 February 2026. Fixed and disclosed, so a small deduction (https://github.com/qdrant/qdrant/security).

## 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.

- The full duration of the 14 August 2026 network-access incident, which the status page only shows as 3 minutes for one region
- How quickly maintainers reply to bug reports, since the issue list didn't show reply counts
- Whether the official MCP server will gain collection management or filtered search

## Weaknesses

- The MCP server has 2 tools, can't manage collections or run filtered queries, and last released on 10 December 2025
- No per-unit rate card for Qdrant Cloud, only a calculator
- Self-hosted builds send usage statistics by default until you opt out
- 474 open issues on the server repo, including a batch of July bug reports
- Keyword search means setting up sparse BM25 vectors, not a plain text query

## 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 payload indexes on the fields you filter on, or filtered search slows on large collections
- Use `/points/query` with `prefetch` to fuse dense and BM25 results
- Pass `wait=true` on upserts when the next step reads its own writes
- On 429, wait for the `Retry-After` seconds before retrying
- Set `with_vector=false` unless the task needs vectors, since they dominate response size

## What the review panel asked for

- MCP tool annotations (2 reviews)
- Programmatic cluster creation
- MCP collection tools
- Key minting by API
- when-not-to text
- filtered search over MCP
- when-not guidance
- Publish Cloud rate limits
- a security.txt
- a deprecation notice period
- Publish per-unit Standard rates

## 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.