Telegram Bot API

# Fix list: Telegram Bot API

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/telegram-bot, the October 2026 research run, assessed 9 October 2026. Grade E, 43 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 Telegram Bot API: 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. Reliability, 40 out of 100, up to 12 more on the total

Why it scored 40: Graded on the hosted API at api.telegram.org, with the hosted lines. No status page was found on core.telegram.org or telegram.org (0). With no readable incident history the record line takes the floor (5). Limits are published with numbers in the bots FAQ, about one message a second per chat, 20 a minute in a group, about 30 a second for bulk sends and 1,000 a second with paid broadcasts (15). Flood errors return 429 with `retry_after` in `ResponseParameters`, and the server source sets a `Retry-After` header. No idempotency key exists for sends (10). No SLA was found, and section 12.3 of the developer terms disclaims uptime and availability (0). The API is generally available at version 10.3 (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.

## 2. Security & auth, 35 out of 100, up to 11.4 more on the total

Why it scored 35: One token per bot gives full control, issued and regenerated in @BotFather with `/token`. It has no scopes or expiry. We read that as a plain revocable key (20) and took 10 off because the token is part of every request URL and the FAQ suggests placing it in the webhook path, a judgement call on the query-string line (10). Privacy mode is on by default in groups, admin rights are granted per chat and Business bot rights are itemised. There is no read-only token or approval step before a send (10). Inbound user messages are untrusted content, and no prompt-injection guidance was found in the reference or features page (3). No audit log of API calls was found. `getWebhookInfo` reports the last delivery error (2). A bug bounty with a safe-harbour statement and security@telegram.org, a webhook secret header and published source IP ranges. No security.txt, SOC 2 or ISO 27001 statement was found (10).

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.

## 3. Schema & documentation, 44 out of 100, up to 9.1 more on the total

Why it scored 44: No OpenAPI document or other machine-readable contract is published by Telegram (0). core.telegram.org/llms.txt answers with a not-found page and no Markdown copy of the docs was found (0). Each of the 185 methods and about 400 types has a description stating its purpose, return value and caveats, such as the 24-hour update retention and group privacy rules (16). Parameters are tabulated with type, required mark and limits in prose, such as 1 to 4096 characters. Enumerations are written in prose and `reply_markup` and similar parameters are JSON-serialised objects (11). Few request examples, and no error catalogue. The reference says `error_code` contents are subject to change (5). Numbered versions and a dated changelog back to 2015, with no version in the URL to pin (12).

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.

## 4. Agent ergonomics, 48 out of 100, up to 8.5 more on the total

Why it scored 48: Responses are whole `Message` or `Update` objects with no field selection. `getUpdates` takes `limit` (1 to 100) and `allowed_updates` to narrow what arrives (12). `getUpdates` pages by `offset`, with `limit`, `timeout` and `allowed_updates`. A bot cannot list or search past messages in a chat (12). Errors carry `ok`, `error_code`, `description` and optional `parameters` with `retry_after` or `migrate_to_chat_id`. Descriptions are free text with no published list (10). No idempotency key for sends. `update_id` lets a receiver discard repeated inbound updates (5). `sendMessage` needs two parameters, GET and POST both work with four parameter encodings, and long polling needs no public server. No official SDK was found (9).

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.

## 5. Payments & pricing, 40 out of 100, up to 7.5 more on the total

Why it scored 40: No x402, MPP or L402 in the reference, the FAQ, the developer terms or the server source (0). The API is free, stated without a login, and the one fee found is 0.1 Stars per broadcast message above 30 a second, for bots with 100,000 Stars and 100,000 monthly users (20). No card or payment method is needed to create a bot and send (20). A person with a Telegram account, which needs a phone number, must create the bot in a chat with @BotFather. Managed bots also need a user to confirm in a Telegram client (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.

## 6. Maintenance & community, 38 out of 100, up to 5.4 more on the total

Why it scored 38: Closed service with an open-source server. The newest dated API change is Bot API 10.3 on 24 August 2026, 46 days before the check (20). Two dated changelog entries fall in the 90 days to 9 October 2026, 10.2 on 14 July and 10.3 on 24 August, so the three-entry line isn't met (0). A changelog, the @BotNews and @BotTalk channels and @BotSupport exist. tdlib/telegram-bot-api has 74 open issues, and we did not read how quickly they are answered (8). No official SDK was found. The official server source is current at 10.3 (5). The server repository had 44 commits between 11 July and 25 August 2026 and has no public CI configuration (5).

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.

## 7. Transparency & trust, 64 out of 100, up to 3.2 more on the total

Made of editorial 53, provenance 74.

Why it scored 64: Editorial half only. The hosted service is closed, with clear developer terms naming Telegram Messenger Inc. The Bot API server source is public under BSL-1.0, an OSI-approved licence (22). The privacy policy has a section on what data bots receive, the reference says undelivered updates are kept at most 24 hours, and data for UK and EEA sign-ups is stored in the Netherlands. No DPA and no retention period for bot messages were found (15). No deprecation policy. Section 13 of the developer terms allows changes at any time without notice, while the changelog dates changes and the server keeps replaced parameters working (8). Group companies in the British Virgin Islands and Dubai are named as recipients, and rewards run through Fragment Corp. No sub-processor list was found (8).

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):

- Terms of service: read, states 5 of the 7 things a reader expects, and has 1 clause that costs points (6.3 of 10)
- Privacy policy: read, states 5 of the 8 things a reader expects (7.8 of 10)
- Status page: not found (0 of 10)
- security.txt: not found (0 of 10)

## 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: whether Telegram runs a public status page on a host not linked from the docs. None was found on core.telegram.org or telegram.org, and it was scored as absent
- unchecked: how quickly issues on tdlib/telegram-bot-api are answered, and the content of bugs.telegram.org, which were not opened
- unchecked: community libraries on core.telegram.org/bots/samples, which was not opened. No official SDK is named in the pages read
- unchecked: the US dollar value of a Telegram Star, so the paid broadcast fee has no unit price
- The developer terms carry no date. Section 4.3 prohibits data collection aimed at large datasets, machine learning models and AI products, and section 5.2(f) prohibits circumventing rate limits. Read both before any probe is run
- The Security score takes 10 off for the token in the URL path. The checklist names the query string, so this is a judgement call
- The changelog says 10.3 replaced `receiver_user_id` and `callback_query_id` 41 days after 10.2 added them. The server source at 10.3 still reads `receiver_user_id`, so no deduction was taken
- telegram.org/privacy redirected to a /gb edition from our network. Other regional editions were not read
- Not tested with a live bot, as we had no token
- The lead held up. The category has no Telegram capability key, so only `messaging.inbound` is recorded

## Weaknesses

- No OpenAPI document, llms.txt or official SDK was found. The contract is one HTML reference page of about 860 KB
- One token gives full control of the bot, sits in the URL path of every request, and has no scopes or read-only form
- No status page, SLA or incident history was found, and the developer terms disclaim uptime and allow changes without notice
- No idempotency key exists for sends, and `error_code` contents are described as subject to change with no published error catalogue
- A person with a Telegram account must create the bot in a chat with @BotFather, and a bot cannot message a user who has not started it
- Developer terms section 4.3 prohibits collecting data for large datasets, machine learning models and AI products. Recorded as a fact, and it matters before any probe is run

## 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.

- Send `POST https://api.telegram.org/bot<token>/sendMessage` with `chat_id` and `text`. Keep the token out of logs, because it is part of the URL.
- On HTTP 429 wait the `parameters.retry_after` seconds before retrying. Stay under one message a second per chat and 20 a minute per group.
- Call `getUpdates` with `offset` set to the last `update_id` plus one, or the same updates return. Updates are kept for at most 24 hours.
- A retried send can post twice, as there is no idempotency key. Record the returned `message_id` before retrying after a timeout.
- In groups the bot sees only commands and replies unless privacy mode is disabled in @BotFather or the bot is an admin.

## 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.