Head to head · Document parsing · October 2026 research run

Docling vs Reducto API + MCP

Reducto API + MCP and Docling score within a point of each other on agent readiness, 65 (B) and 64.3 (B). Docling leads on reliability, security & auth, payments & pricing, maintenance & community and transparency & trust. Both do document parsing.

Best document parsing, OCR and extraction APIs for AI agents · All 182 documents comparisons

Which one, for what

Docling B

Best for Self-hosted PDF and Office conversion to Markdown or JSON for RAG and agents that can't send documents out.

Ahead on

  • Reliability, 85 against 77
  • Security & auth, 53 against 40
  • Payments & pricing, 60 against 30
  • Maintenance & community, 89 against 80
  • Transparency & trust, 65 against 59

Also in its favour

  • No key needed to call it
  • Open source

Watch for

28 advisories from 2 June to 23 October 2026, five High, including arbitrary file read and write through the opt-in Tectonic TikZ engine

Reducto API + MCP B

Best for Messy PDFs, tables and spreadsheets where an agent needs parse, extract and split behind one small MCP.

Ahead on

  • Agent ergonomics, 84 against 77

Also in its favour

  • A hosted endpoint, with nothing to install
  • No incidents deducted, where Docling loses 8 points for them

Watch for

Zero data retention, the no-training promise and DELETE endpoints only on Growth and above; Standard retention not stated

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Payments & pricing Docling +30

Docling 60

Self-hosted rule. Free MIT software with nothing to buy, so 20 + 20 + 20, and no payment protocol (0). 60.

Reducto API + MCP 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).

Security & auth Docling +13

Docling 53

Framework reading. No credentials for local use; docling-serve has one optional shared key, off by default (20 of 30). Remote services are off unless explicitly enabled, local file fetch is guarded by flags, and MCP tools carry read-only and destructive annotations; manipulation tools can delete items and save files (12 of 20). Converted documents are untrusted text and no prompt-injection guidance was found (3 of 15). Logging only, no audit trail (3 of 15). A disclosure policy with private reporting, 28 advisories published in public with fixed versions, under the LF AI & Data Foundation; no bug bounty (15 of 20). 20+12+3+3+15 = 53. The advisories themselves are the deduction.

Reducto API + MCP 40

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 is valid (support@reducto.ai, expires 1 July 2027), the SDK repositories carry a SECURITY.md with a reporting address, and the data policies page says SOC 2 Type I and Type II are complete with the report on request, with HIPAA processing on Growth and Enterprise and regular penetration testing. No ISO 27001, bug bounty or public advisories found. That sits above pandadoc (8, a SOC 2 claim and public bulletins) and below oxylabs-web-scraper (14, ISO 27001 as well) (10 of 20).

Maintenance & community Docling +9

Docling 89

v2.148.0 on 5 November 2026, the day before this check (30). 13 releases from 9 October to 5 November (20). Open issues and pull requests are numerous, an AI triage workflow and PR reminders run, and six maintainers are named; replies not sampled (17 of 25). docling-mcp publishes to the official MCP registry on release under io.github.docling-project, per its workflow and server.json; the registry entry itself was not read (12 of 15). CI, a dependency-update workflow and uv.lock (10). 30+20+17+12+10 = 89.

Reducto API + MCP 80

Python SDK v0.25.0, Node.js SDK v0.19.0 and Go SDK v0.2.0, all on 24 October 2026 (30). In the 90 days from 4 August, Python v0.23.0 (3 September), v0.24.0 (8 September) and v0.25.0, Node.js v0.17.0 (3 September), v0.18.0 (10 September) and v0.19.0, and Go v0.1.0 (10 September) and v0.2.0, eight SDK tags counted half (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). The SDK repositories run CI, end-to-end and spec-drift workflows. The MCP server's repository, read on 1 October, no longer answers a clone (10).

Reliability Docling +8

Docling 85

Local-software lines, graded on the library with docling-serve and docling-mcp. Installs from PyPI (docling 2.148.0, per the PyPI release feed) for Python 3.10 and later on macOS, Linux and Windows (20). Public CI workflows for checks, tests and heavy examples; the pass state of the default branch was not read (20 of 25). About 719 open issues and 223 open pull requests on the repository page against 13 releases in four weeks, many of them fixes; replies not sampled (15 of 25). Semantic-release changelog with Feature and Fix sections and Breaking sections where they apply (15). Version 2.x since 2024 (15). 20+20+15+15+15 = 85.

Reducto API + MCP 77

incident.io status page at status.reducto.ai with API, Parsing, Extraction, Splitting, Editing and Studio components (20). Its feed lists nine incidents from 13 August to 1 October 2026 and none since, each a latency rise or partial degradation of one product (Extract on 13 August, Parse on 13 August, Parse and Extract with citations on 18 August, Deep Split for about 11 minutes on 9 September, Parse with embed_pdf_metadata on 16 September, Parse on 28 September, and Parse, then Parse in the EU region, on 1 October). The history page answered 500 on 2 November, so 4 to 12 August rests on the 1 October reading, which found only partial incidents after 23 July. All partial, read as minor (20 of 30). 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).

Agent ergonomics Reducto API + MCP +7

Docling 77

About 25 MCP tools in the default groups (15), selectable as conversion, generation and manipulation groups, with anchor tools to read one part of a document (22 of 25). max_num_pages, max_file_size and page ranges on conversion, Markdown or JSON export (15 of 20). Errors are exceptions with messages, such as an unknown document key listing the valid ones (12 of 20). readOnlyHint and destructiveHint on the MCP tools, and conversions are repeatable (18 of 20). Sensible defaults and one call to convert; Python only, with a docling-client package for docling-serve (10 of 15). 22+15+12+18+10 = 77.

Reducto API + MCP 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).

Transparency & trust Docling +6

Docling 65

MIT, an OSI licence, with model licences noted as separate (30). Runs locally, remote services off by default and air-gapped use documented; no privacy statement is needed for local use, but model downloads from Hugging Face on first run are only partly described (22 of 30). Breaking changes are marked in the changelog, with no deprecation policy or dated notices (10 of 20). No telemetry found in the source; nothing to opt out of (15 of 20). 30+22+10+15 = 77.

Reducto API + MCP 59

Closed service under published terms (effective 17 April 2026). The MCP server's MIT repository no longer answers a clone (15). The data policies page gives zero data retention on Growth and above, purging API-submitted artifacts within 24 hours except Studio jobs and persist_results, and says Growth and above data is never used for training. Standard gets neither promise, the deletion docs give no retention period for it, and the terms grant Reducto a licence to use uploaded content to develop and improve the Services and for its other business purposes. The privacy policy dates from 21 January 2024 (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 docs with REDUCTO_TELEMETRY=0 to opt out. Subprocessors sit in the JavaScript-only trust centre, with EU and AU endpoints on Growth and an EU data residency page (8 of 20).

Score by category

CategoryWeight this runDoclingReducto API + MCPEdge
Reliability16%208577Docling +8
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.27780Reducto API + MCP +3
Agent ergonomics13%16.27784Reducto API + MCP +7
Security & auth14%17.55340Docling +13
Payments & pricing10%12.56030Docling +30
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.88980Docling +9
Transparency & trust7%8.86559Docling +6
Negative events≤15-80
TotalB 64.3/100B 65/100

Facts side by side

FactDoclingReducto API + MCP
KindSDK + MCPHTTP API
VendorDocling Project (LF AI & Data Foundation, started at IBM Research Zurich)Reducto
Hosted endpointno (local only)https://platform.reducto.ai
Transportsstdio, Streamable HTTP, HTTPHTTP, Streamable HTTP, stdio
AuthNoneAPI key
PricingFreePay per use
Price for document parsingfree$10 per 1,000 pages
x402nono
LicenceMITnone
Tools exposed259
Read-only variant documentednono
llms.txtnoyes
MCP registryio.github.docling-project/docling-mcpnot listed
Last release2026-11-052026-10-24
Terms last updatedno document linked2026-04-17
Privacy policy last updatedno document linked2024-01-21
Customer content may train modelsnot read yetnot found in the text
Terms restrict automated accessnot read yetyes
Terms restrict benchmarkingnot read yetyes
Terms or service can change without noticenot read yetnot found in the text
Arbitration or class-action waivernot read yetyes
Popularity70k stars, 947k PyPI/wk342k npm/wk, 1.1M PyPI/wk
Agent reviewsnone4/5 (2)

How the scores have moved

Verdicts

Docling

An MIT Python library with a CLI, a REST server and a 25-tool MCP server converts PDFs, Office files, HTML and scans into structured documents on the user's own machine, with near-daily releases. Twenty-eight security advisories were published between June and October 2026, five rated High, all for untrusted-input handling.

Reducto API + MCP

Nine MCP tools with when-to-use descriptions, jobid:// chaining and SDKs in three languages released on 24 October 2026. Zero data retention, a no-training promise and DELETE endpoints are for Growth and above only, while Standard's terms licence uploaded content to improve the service. Keys carry no scopes.

Before you call either

Docling

  1. Keep enable_remote_services and remote image fetching off when converting untrusted files, and set max_num_pages and max_file_size
  2. Upgrade to v2.148.0 or later before parsing untrusted documents, since recent releases fix memory-exhaustion and file-read flaws
  3. Start docling-mcp with only the tool groups the task needs, and use the anchor tools to read parts of a long document
  4. Treat converted text as untrusted input, never as instructions
  5. Set DOCLING_SERVE_API_KEY before exposing docling-serve beyond localhost

Reducto API + MCP

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

Questions

Which is better for AI agents, Docling or Reducto API + MCP?

Reducto API + MCP and Docling score within a point of each other on agent readiness, 65 (B) and 64.3 (B). Docling leads on reliability, security & auth, payments & pricing, maintenance & community and transparency & trust.

Which is cheaper for document parsing, Docling or Reducto API + MCP?

Docling, at free against $10 per 1,000 pages for Reducto API + MCP. These are the vendors' published prices for the job.

Can an agent call Docling and Reducto API + MCP without installing anything?

Docling runs on your own machine, with no hosted endpoint listed. Reducto API + MCP has a hosted endpoint at https://platform.reducto.ai.

Are Docling and Reducto API + MCP open source?

Docling is open source (MIT). No open-source release is listed for Reducto API + MCP.

Other comparisons with Docling or Reducto API + MCP

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.