Reducto API + MCP by Reducto

# Fix list: Reducto API + MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/reducto, the October 2026 research run, assessed 1 October 2026. Grade B, 63.2 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 Reducto 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, 35 out of 100, up to 11.4 more on the total

Why it scored 35: Plain bearer API key from Reducto Studio, used for the hosted MCP too. The local server adds a device-code login (20 of 30). No read-only or scoped key, and five of the nine tools start billable jobs with no confirmation step (5 of 20). Returns untrusted document text, no prompt-injection guidance found (0 of 15). get_job and list_jobs let the operator review past jobs, no audit log found (5 of 15). security.txt valid per the provenance check. The trust centre at trust.reducto.ai renders no certifications without JavaScript, and security contact goes to support@reducto.ai (5 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.

## 2. Payments & pricing, 30 out of 100, up to 8.8 more on the total

Why it scored 30: No x402, MPP or L402 (0). Per-1,000-page list prices for every endpoint on the public pricing page (20). $150 of free usage (15,000 credits) on Standard, and the page doesn't say whether a card is needed (10 of 20, our call until that's confirmed). A person signs up in a browser, or completes a device-code login for the local MCP (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.

## 3. Reliability, 72 out of 100, up to 5.6 more on the total

Why it scored 72: incident.io status page at status.reducto.ai with API, Parsing, Extraction, Splitting, Editing and Studio components and a history page (20). The history lists 11 incidents since 23 July, nearly all latency or partial degradation of Parse, Extract or Deep Split. The exception is elevated failures on OCR-dependent Parse and Extract on 23 July, during an upstream Google Cloud Vision degradation. Most durations aren't stated, and two Parse degradations were open on 1 October (15 of 30, our call). Edge limits of 1,000 requests a second per key and 300 on job polling (15). 429 bodies carry a code and say to back off or use webhooks, and the SDKs retry 429s with exponential backoff. No Retry-After and no idempotency key (12 of 15). Custom SLA on Enterprise only (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.

## 4. Transparency & trust, 60 out of 100, up to 3.5 more on the total

Made of editorial 38, provenance 82.

Why it scored 60: Closed service under published terms, MCP server MIT (15). Zero data retention purges API artifacts within about 24 hours, and DELETE endpoints exist, but both are Growth and Enterprise only. Standard's default retention and any training use aren't stated in the docs we read (10 of 30). The 1 September pricing change has a dated migration page, and we found no deprecation policy (5 of 20). The MCP server sends PostHog usage telemetry by default, disclosed in the README with REDUCTO_TELEMETRY=0 to opt out. Hosted subprocessors and data locations sit in the JavaScript-only trust centre, with EU and AU endpoints on Growth (8 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: reducto.ai, registered 2023-09-23 (3 years) (7 of 15)
- Changelog: not found (0 of 10)

## 5. Schema & documentation, 80 out of 100, up to 3.3 more on the total

Why it scored 80: OpenAPI spec at docs.reducto.ai/openapi.json per the 30 September check, and the SDK repositories run a spec-drift workflow against it (25). llms.txt per the same check (10). Each MCP tool description says when to use it and how to chain results through jobid:// and get_job, though none says when not to (17 of 20). MCP parameters are plain strings checked against enums at run time, and options takes a free-form dict or JSON string (8 of 15). Validation errors carry a 'What to do' line, and 429 codes 1000 and 2000 are documented (12 of 15). SDK changelogs live on GitHub, but the docs changelog is the password-protected on-prem one. The 1 September pricing migration has a dated page (8 of 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.

## 6. Agent ergonomics, 84 out of 100, up to 2.6 more on the total

Why it scored 84: Nine MCP tools with short descriptions (25). page_range, chunk_mode, return_images, URL results for large outputs, and list_jobs (18 of 20). Errors say what to do next, and 429 bodies name the fix (18 of 20). jobid:// reuse avoids paying to parse twice, but there's no idempotency key and no readOnlyHint or destructiveHint annotations (8 of 20). parse_document needs only document_url. SDKs in Python, Node.js and Go (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.

## 7. Maintenance & community, 80 out of 100, up to 1.8 more on the total

Why it scored 80: Node.js SDK v0.18.0 on 2026-09-10 and Python SDK v0.24.0 on 2026-09-08, and the MCP server moved to MCP Python SDK 2.x on 2026-09-24 (30). Python v0.23.0 and v0.24.0 and Node.js v0.17.0 and v0.18.0 in September (20). No public changelog for the hosted API and issue replies not sampled (5 of 15). Current SDKs in Python, Node.js and Go (15). SDK repositories run CI, end-to-end and spec-drift workflows, and the MCP server tests on Python 3.11 to 3.13 (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: Standard plan default retention and whether customer documents train models
- unchecked: certifications and subprocessors in the JavaScript-only trust centre
- unchecked: whether the $150 free usage needs a card
- The listing said the MCP server has six tools. The docs and source list nine. Patched toolCount, notable and the MCP detail

## Weaknesses

- Zero data retention and DELETE endpoints only on Growth and above, Standard retention not stated
- No public changelog for the hosted API
- 11 incidents since 23 July, mostly latency, with Parse degraded on 1 October
- MCP server sends usage telemetry by default
- Plain API keys with no scopes, and no annotations on tools that start billable jobs

## 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.

- Use `/parse_async` or `/extract_async` with a webhook for anything long; sync calls time out at 900 seconds
- Pass `jobid://<job_id>` from a parse into extract or split so you don't pay to parse twice
- Call `get_job` when a tool returns a URL result instead of reading fields from the truncated reply
- The hosted MCP can't read local files; run `uvx mcp-server-reducto` when the agent needs to upload from disk
- Set `REDUCTO_TELEMETRY=0` in the local server's environment to turn off usage telemetry

## What the review panel asked for

- Enums in the schema
- Annotate billable tools
- public API changelog

## 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.