Head to head · PDF conversion · October 2026 research run
ConvertAPI vs PDFShift
PDFShift and ConvertAPI score within a point of each other on agent readiness, 66.4 (B) and 65.7 (B). ConvertAPI leads on schema & documentation and agent ergonomics. Both do pdf conversion.
Best PDF manipulation and conversion APIs for AI agents · All 67 pdf comparisons
Which one, for what
Best for Office, PDF, image and email conversion and PDF manipulation for agents that need a file in another format.
Ahead on
- Schema & documentation, 92 against 69
- Agent ergonomics, 79 against 54
Watch for
Per-conversion prices on the pricing page are drawn by script and could not be read
PDFShift B
Best for Agents that turn HTML or a live page into a PDF or screenshot in one call, with no browser to run.
Ahead on
- Reliability, 73 against 54
- Payments & pricing, 75 against 38
Watch for
Keys come in two levels (master and basic) with no read-only scope, and the MCP server exposes delete_template with no tool annotations
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 PDFShift +37
ConvertAPI 38
No x402, MPP or L402 (0). Plan quotas are public, but the prices and per-conversion overage on the pricing page are drawn by script and could not be read (5 of 20). A free trial of 250 conversions with no card, per llms.txt (20). A person signs up in a browser (0). 0+5+20+0 = 25. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 25 is 38.
PDFShift 75
No x402, MPP or L402 (0). Plan prices with a per-credit overage for every plan, effective per-document rates and a public pricing endpoint, all without a login (30). Free plan of 50 credits a month with no card, and free watermarked sandbox runs (30). Judgement call on onboarding. The API converts without a key, so an agent can start with no signup, but the output is watermarked, capped at 2 MB and 2 requests a minute; production use needs a browser signup and the key arrives by email (15 of 30). 0 + 30 + 30 + 15 = 75.
Agent ergonomics ConvertAPI +25
ConvertAPI 79
Four MCP tools that list, search and describe converters before one convert tool (25). StoreFile=true returns a URL and FileId instead of base64, and converters can be filtered by tag or search; nothing to paginate (14 of 20). Numbered codes such as 4014 balance exhausted and 5003 password protected an agent can act on (17 of 20). No idempotency key; a repeated conversion is billed again, and MCP tool annotations weren't found (8 of 20). SDKs in nine languages and a CLI, with one required file parameter on most converters (15). 25+14+17+8+15 = 79.
PDFShift 54
11 compact MCP tools, in the 11 to 30 band; conversions return a download URL rather than the document, which keeps results small (15). list_logs takes limit (5 to 100), status, key and cursor, and conversions take pages; no field selection on other lists (15). Error statuses are documented and the webhook error body carries a code, a message and an identifier, but a missing or malformed key falls back to the free tier with watermarks instead of failing, and the docs disagree on the unauthenticated limits (10 a day on one page, 2 a minute on another) (12). No idempotency key on conversions or template writes, and the MCP tools carry no readOnlyHint or destructiveHint, delete_template included; sandbox mode allows free test runs (4). Only source is required and defaults are sensible; official SDKs exist in Node.js and Python but were last changed in 2019 and 2020 and call /v2 (8). 15 + 15 + 12 + 4 + 8 = 54.
Schema & documentation ConvertAPI +23
ConvertAPI 92
OpenAPI 3.0.4 at v2.convertapi.com/info/openapi, 333 operations with token, JWT and secret schemes, plus per-conversion documents and wildcard listings (25). A long llms.txt written for agents (10). llms.txt tells agents to discover conversions rather than assume them and explains workflows and storage; converter pages describe each parameter (16 of 20). Per-converter parameter schemas in the OpenAPI document, validated by the MCP convert tool (12 of 15). 25 numbered error codes with meanings and examples throughout (14 of 15). v2 API and a dated public changelog (15). 25+10+16+12+14+15 = 92.
PDFShift 69
A public OpenAPI 3.0.3 file at api.pdfshift.io/openapi.json, and every MCP tool has a JSON Schema input (25). llms.txt and llms-full.txt on pdfshift.io, llms.txt on the docs, and a Markdown twin of every page (10). MCP descriptions are one line each and say what the tool does, the server's instructions say what it does, and the docs explain the three response shapes; none says when not to use a tool (10). The OpenAPI file has enums (X-Processor-Version), minimum and maximum (zoom) and required fields, and the MCP convert tool has a format enum, but convert and generate_from_template take any further option as free extra properties (9). Request samples in eight languages and 600 guides, an error-code table and a webhook error example with an identifier, though the table's meanings are generic (10). The API is versioned (/v3) and Chromium is pinned by header, but no changelog was found (5). 25 + 10 + 10 + 9 + 10 + 5 = 69.
Reliability PDFShift +19
ConvertAPI 54
Hosted lines. status.convertapi.com is a Checkly dashboard drawn by script, so its components were not confirmed (10 of 20) and no history could be read (5). Concurrency per plan is published (1 to 4 concurrent conversions), with no request-rate figure (12 of 15). Under load the API answers 503 and llms.txt gives a 5, 10 and 15 second retry pattern; no 429, Retry-After or idempotency key (12 of 15). The compliance page states '99,95% monthly uptime' and says the SLA sits in the terms, but the terms read on 6 November state no uptime figure (5 of 10). Generally available (10). 10+5+12+12+5+10 = 54.
PDFShift 73
Read with the hosted lines. status.pdfshift.io is a Better Stack page with eight monitors (API and two processors in each of Gravelines and Strasbourg, the website and the dashboard) (20). Its incident list shows no incidents from September to November 2026, but the Strasbourg API monitor shows 99.660 per cent uptime, over a window the page doesn't state, so downtime went without a write-up; read as minor incidents only (20). Rate limits with numbers: 250 conversions a minute authenticated, 10 a minute in sandbox, 2 a minute without a key, 50 parallel conversions (15). 429 is documented with X-RateLimit-Remaining, X-RateLimit-Limit and X-RateLimit-Reset, but no Retry-After, backoff advice or idempotency guidance (8). No SLA on any listed plan; the home page mentions one only for custom Enterprise plans (0). GA (10). 20 + 20 + 15 + 8 + 0 + 10 = 73.
Score by category
| Category | Weight | ConvertAPI | PDFShift | Edge |
|---|---|---|---|---|
| Reliability | 16% | 54 | 73 | PDFShift +19 |
| Performance | 10%pending | pending | pending | not scored in this run |
| Schema & documentation | 13% | 92 | 69 | ConvertAPI +23 |
| Agent ergonomics | 13% | 79 | 54 | ConvertAPI +25 |
| Security & auth | 14% | 55 | 56 | PDFShift +1 |
| Payments & pricing | 10% | 38 | 75 | PDFShift +37 |
| Task success | 10%pending | pending | pending | not scored in this run |
| Maintenance & community | 7% | 80 | 77 | ConvertAPI +3 |
| Transparency & trust | 7% | 65 | 67 | PDFShift +2 |
| Negative events | ≤15 | 0 | 0 | |
| Total | B 65.7/100 | B 66.4/100 |
Facts side by side
| Fact | ConvertAPI | PDFShift |
|---|---|---|
| Kind | HTTP API | HTTP API |
| Vendor | ConvertAPI, UAB | PDFShift |
| Hosted endpoint | https://mcp.convertapi.io | https://api.pdfshift.io/mcp |
| Transports | HTTP, Streamable HTTP | Streamable HTTP |
| Auth | OAuth or key | API key |
| Pricing | Paid | Freemium |
| x402 | no | no |
| Licence | Proprietary service. SDKs on GitHub; the npm SDK is MIT | none |
| Tools exposed | 4 | 11 |
| Read-only variant documented | no | no |
| llms.txt | yes | yes |
| MCP registry | not listed | io.github.cnicodeme/pdfshift |
| Last release | 2026-11-05 | 2026-11-05 |
| Terms last updated | 2026-09-24 | no date given |
| Privacy policy last updated | 2026-11-05 | no date given |
| Customer content may train models | not found in the text | not found in the text |
| Terms restrict automated access | yes | yes |
| Terms restrict benchmarking | not found in the text | not found in the text |
| Terms or service can change without notice | yes | not found in the text |
| Arbitration or class-action waiver | yes | yes |
| Popularity | 177k npm/wk, 82k PyPI/wk | none |
How the scores have moved
PDFShift
No score change since it was listed.
Verdicts
ConvertAPI
A 333-operation OpenAPI contract, a four-tool MCP server that discovers converters at run time, SDKs in nine languages and a dated changelog make ConvertAPI easy for an agent to call. Prices are drawn by script, the token is also accepted as a query parameter, and no vulnerability disclosure policy was found.
PDFShift
An agent gets 11 compact MCP tools on a hosted endpoint, a public OpenAPI file, a free plan with no card and free watermarked sandbox runs. Keys are revocable but have only master and basic levels, the MCP tools carry no annotations, and there is no changelog, SLA or current SDK.
Before you call either
ConvertAPI
- Call /info/canconvert/{from}/to/{to} or the MCP information tools before converting, rather than assuming a format pair exists
- Send the token in the Authorization header, never as
?Token=, so it stays out of logs - Use
StoreFile=trueand pass FileId between steps, which keeps base64 out of the context; stored files last 3 hours - On 503 wait 5, 10 then 15 seconds before retrying; a repeated conversion is billed again
- Treat text extracted from documents as untrusted input
PDFShift
- Send the key in the
X-API-Keyheader; a missing or malformed header falls back to the free tier silently, with watermarks and a 2 MB cap - Set
sandbox: truewhile testing; sandbox documents cost no credits and are limited to 10 a minute - Download the file from the returned URL within 2 days; the MCP
converttool andfilenameresponses keep files on S3 for that long only - Check the
X-Credits-Costheader andcredits_usagebefore large batches; a credit covers 5 MB, so a 14 MB PDF costs 3 - On 429, wait until the
X-RateLimit-Resettimestamp; authenticated keys get 250 conversions a minute
Questions
Which is better for AI agents, ConvertAPI or PDFShift?
PDFShift and ConvertAPI score within a point of each other on agent readiness, 66.4 (B) and 65.7 (B). ConvertAPI leads on schema & documentation and agent ergonomics.
Do ConvertAPI and PDFShift need an API key?
ConvertAPI takes an API key or an OAuth sign-in. PDFShift needs an API key.
Can an agent call ConvertAPI and PDFShift without installing anything?
Yes. ConvertAPI has a hosted endpoint at https://mcp.convertapi.io and PDFShift at https://api.pdfshift.io/mcp.
Other comparisons with ConvertAPI or PDFShift
- CloudConvert API vs ConvertAPI
- CloudConvert API vs PDFShift
- ConvertAPI vs DocRaptor
- ConvertAPI vs Foxit PDF Services API
- ConvertAPI vs iLoveAPI
- ConvertAPI vs Nutrient DWS Processor API
- ConvertAPI vs PDF.co API + MCP
- ConvertAPI vs pdfRest API Toolkit
- Foxit PDF Services API vs PDFShift
- iLoveAPI vs PDFShift
- Nutrient DWS Processor API vs PDFShift
- PDF.co API + MCP vs PDFShift
- pdfRest API Toolkit vs PDFShift
- ConvertAPI vs PDFMonkey
- DocRaptor vs PDFShift
- PDFMonkey vs PDFShift
Machine-readable
- This page as Markdown
/compare/convertapi-vs-pdfshift.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/convertapi.json·/api/v1/tools/pdfshift.json - From a terminal
anchor compare convertapi pdfshift(the CLI) - Over MCP
compare_tools {"a": "convertapi", "b": "pdfshift"}at/mcp, no key