Jina Reader by Jina AI (Elastic)

# Fix list: Jina Reader

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/jina-reader, the October 2026 research run, assessed 1 October 2026. Grade E, 45.3 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 Jina Reader: 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, 32 out of 100, up to 11.9 more on the total

Why it scored 32: One Bearer key per account, revocable, no scopes, shared with Search, Embeddings and Reranker, and Reader also runs with no key at all (20). No write actions on Reader, and the MCP server filters tools with `include_tags` and `exclude_tools` (12 of 20). Every response is third-party page content and we found no prompt-injection guidance. `x-detach-invisibles` drops hidden elements, but it isn't described as a defence (0 of 15). No per-request log found (0 of 15). No security.txt, the privacy statement is Elastic's, and two SSRF reports (#1252 and #1253, 12 June 2026) sit open as public issues with no reply we could see and no commit since 22 May (0 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. Schema & documentation, 43 out of 100, up to 9.3 more on the total

Why it scored 43: No OpenAPI file found (0 of 25). jina.ai/llms.txt returned 404 in last week's check and issue #1224 asking for one is open, but the GitHub README is a full Markdown reference of every header (5 of 10). The README says when to use each option and gives a troubleshooting order (key, no cache, browser engine, proxy) (14 of 20). Options are headers with documented values, such as `x-engine` browser, curl or auto, and the MCP tools have typed schemas (11 of 15). curl examples and a cookbook, but no documented error responses (7 of 15). No API changelog. The README carries a dated updates list, last April 2026 (6 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.

## 3. Payments & pricing, 35 out of 100, up to 8.1 more on the total

Why it scored 35: No x402, MPP or L402 (0). Billed in tokens with no price in any currency on the Reader page, which points to a separate table (0). Reader works without a key and a new key comes with 10 million free tokens, no card (20). An agent can read pages with no signup at 20 a minute, but Search needs a key (15 of 20).

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.

## 4. Reliability, 60 out of 100, up to 8 more on the total

Why it scored 60: Atlassian Statuspage at status.jina.ai with automated incidents per service (20). The feed only reaches back to 15 September 2026. In that window r.jina.ai had no incidents and s.jina.ai had eight five-minute ones, seven with error rates of 8 to 13 per cent and one at 45 per cent on 15 September. Minor only, less 5 for the two-week window (15 of 30). Rate limits published, Reader 20 a minute without a key, 500 with a free or paid key and 5,000 premium, Search 100 and 1,000 (15). No 429 or Retry-After guidance found on the Reader page or in the README (0). No SLA found (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.

## 5. Agent ergonomics, 76 out of 100, up to 3.9 more on the total

Why it scored 76: `x-max-tokens` truncates and `x-token-budget` rejects oversized pages, `x-target-selector` returns one element, and images, links and media can be dropped or reduced to text (25). Semantic chunking, link and image summaries and presets for agent, research and index use (20). No documented error codes (8 of 20). Reads are GETs and safe to retry, with a cache and `x-no-cache` to bypass it. The MCP tools carry no readOnly or destructive hints (13 of 20). No required options beyond the URL, but no official SDK, only HTTP and the MCP server (10 of 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, 61 out of 100, up to 3.4 more on the total

Made of editorial 45, provenance 76.

Why it scored 61: The Reader repository is the Apache-2.0 open-source branch of the code behind r.jina.ai, minus the storage layer (25 of 30). The Reader page says requests are never used for training, and the terms say Jina doesn't store input or output per the 26 September check, but the page gives a 5-minute cache and the README a 3,600-second cache lifetime, and the privacy statement is Elastic's generic one (15 of 30). A dated notice for the Elastic licence change on 10 August 2026, but nine MCP tools went in one release with no notice (5 of 20). No subprocessor list or data location found (0 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: jina.ai, registered 2020-01-20 (6 years) (11 of 15)
- Changelog: not found (0 of 10)
- security.txt: not found (0 of 10)

## 7. Maintenance & community, 65 out of 100, up to 3.1 more on the total

Why it scored 65: The MCP server that fronts Reader reached 1.10.0 on 2026-09-18, after adding OCR and page options to read_url on 15 September. The Reader repository's last commit is 22 May 2026 (30). MCP versions 1.5.0 to 1.10.0 between 11 August and 18 September (20). 22 open Reader issues, including the two SSRF reports from June and a token-usage spike from April, with no fix in the repository (5 of 25). No official SDK, the MCP server is the maintained client (5 of 15). Reader has CI workflows, the MCP repository has none (5 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.

## Deductions

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

- -4: the hosted MCP server's show_api_key tool returned the account's bearer token into the conversation, next to read_url, which fetches untrusted pages, from August 2025 until v1.10.0 removed it on 2026-09-18. Fixed and documented in the README, so deducted less (https://github.com/jina-ai/MCP#functions-have-been-removed-in-v1100).
- -3: v1.10.0 on 2026-09-18 removed nine tools from mcp.jina.ai, including search_web_deep added on 12 August and the parallel_* tools, with no advance notice (https://github.com/jina-ai/MCP/pull/36).

## 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.

- Whether SSRF reports #1252 and #1253 affect hosted r.jina.ai or only self-hosted copies.
- Incident history before 15 September 2026, which the feed doesn't reach.
- Which cache lifetime applies on the hosted Reader, 5 minutes per the product page or 3,600 seconds per the README.
- The price per token, which the public page doesn't show.

## Weaknesses

- No price per token in any currency on the public page
- No OpenAPI, no llms.txt and no documented error responses
- Two SSRF reports open as public issues since 12 June 2026
- Hosted MCP exposed the bearer token through show_api_key until 18 September 2026
- The Reader page and the README disagree on cache lifetime, 5 minutes against 3,600 seconds

## 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.

- Prefix the URL, `https://r.jina.ai/https://example.com`, and read the Markdown
- Send `x-max-tokens` to cap a page, or `x-token-budget` to refuse one that's too big
- Use `x-target-selector` when you need one part of a page
- Keyless is 20 requests a minute and can't search. Add a key for loops or s.jina.ai
- Send `x-no-cache: true` when a page may have changed or a blocked response got cached

## What the review panel asked for

- publish a price per token
- show dollar cost per Search request
- one cache lifetime
- an error reference

## 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.