Runpod by Runpod

# Fix list: Runpod

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/runpod, the October 2026 research run, assessed 1 October 2026. Grade D, 53.7 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 Runpod: 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, 35 out of 100, up to 13 more on the total

Why it scored 35: Better Stack page at uptime.runpod.io with per-service and per-region history (20). Several major incidents in the last 90 days. US-TX-3 network storage was down for about 10 hours on 8 and 9 July, US-IL-1 lost power for 6 hours 10 minutes on 14 and 15 August, and EUR-IS-1 and US-NC-2 had network problems lasting most of a day in August and September; the serverless API and pod proxy also ran elevated errors for 1 hour 25 minutes on 6 September (0). No request rate limits found in the operation reference or the REST v2 overview (0). No 429 or backoff guidance found; `/retry` requeues a failed job and `/cancel` stops one (5 of 15). No SLA found (0). Serverless and REST v2 are GA; batch jobs are beta (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. Payments & pricing, 20 out of 100, up to 10 more on the total

Why it scored 20: No machine payment protocol (0). Per-second serverless and per-hour pod prices published without a login (20). No free tier; prepaid credit only (0). Signup is a browser flow; the MCP's OAuth still needs an account (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. Agent ergonomics, 47 out of 100, up to 8.6 more on the total

Why it scored 47: Results can be streamed in chunks of up to 1 MB and polled by job ID; the MCP server's tool count isn't published (12). Pagination and filtering on REST v2 not found in what we read (5). No error codes or messages documented for serverless operations (5). `/retry`, `/cancel` and `/purge-queue` give safe recovery for jobs; no idempotency keys and no MCP tool annotations found (10). FlashBoot on by default, Python and JavaScript SDKs, the Flash SDK and an official MCP server (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.

## 4. Security & auth, 60 out of 100, up to 7 more on the total

Why it scored 60: API keys can be All, Read Only or Restricted per serverless endpoint, and can be disabled or deleted; the hosted MCP signs in with OAuth. The GraphQL docs still show the key as `?api_key=` in the URL, which costs 10 (20). Read-only keys and per-endpoint restrictions; no confirmation for destructive calls found (15). Runs your own containers and returns their output (10). No audit log found; per-job status and worker logs only (5). security.txt valid until 31 January 2027, and the trust centre lists SOC 2 Type 2, SOC 3, ISO 27001 and HIPAA; no bug bounty or public advisories 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.

## 5. Schema & documentation, 81 out of 100, up to 3.1 more on the total

Why it scored 81: OpenAPI for REST v2 at api.runpod.io/v2/openapi.json, with v1 and GraphQL schemas also published (25). llms.txt with .md pages (10). Guides explain queue-based against load-balanced endpoints and sync against async calls, with result retention for each (14). Job statuses are a fixed set (`IN_QUEUE`, `IN_PROGRESS`, `COMPLETED`, `FAILED`) and payload limits are stated (20 MB sync, 10 MB async, 1 MB per stream chunk) (10). Examples throughout; no error responses documented for the serverless operations (7). Dated release notes and a versioned API (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. Transparency & trust, 65 out of 100, up to 3.1 more on the total

Made of editorial 60, provenance 70.

Why it scored 65: Closed service under Delaware terms; the Python SDK is MIT (20). Sync results are kept 1 minute and async 30 minutes, and the docs claim GDPR compliance for EU regions, but no retention periods for pod disks or volumes after deletion and the DPA sits behind trust-centre access (15). Dated notices, REST v1 retiring on 15 November 2026 (announced 18 August) and GraphQL in early 2027; one model endpoint got two weeks' notice (15). 31 regions named with data-centre codes; subprocessor list not found (10).

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: runpod.io, no registry record we could read (0 of 15)
- Endpoint on the vendor's domain: api.runpod.ai is not on runpod.io (0 of 15)

## 7. Maintenance & community, 82 out of 100, up to 1.6 more on the total

Why it scored 82: Latest release-notes entry 15 September 2026 (30). Thirteen dated entries between 1 July and 15 September (20). Closed service with dated release notes and support; we didn't review GitHub issue response times (12 of 25). Python and JavaScript SDKs, the Flash SDK and the MCP server are current (15). Package health not checked (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.

## 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.

- We couldn't find request rate limits for serverless or REST v2; they may sit in the OpenAPI file, which we didn't read in full.
- The MCP server's tool count and annotations aren't on the docs page; the source is at github.com/runpod/runpod-mcp.
- The previous listing gave the US-IL-1 outage as 13 and 14 August for 6 hours 15 minutes; the incident page says 14 and 15 August, 6 hours 10 minutes.

## Weaknesses

- Data-centre outages of 6 to 24 hours in each of July, August and September 2026
- No published rate limits, 429 guidance, error codes or SLA
- The GraphQL API, live until early 2027, takes the API key in the URL
- No free tier, prepaid credit only
- Three APIs in flight, with REST v1 retiring on 15 November 2026

## 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.

- Create a Restricted or Read Only key per endpoint for the agent, not an All key
- Use REST v2 at api.runpod.io/v2 with a Bearer header; avoid GraphQL, which puts the key in the URL
- Fetch `/run` results within 30 minutes and `/runsync` results within 1 minute, or they're gone
- Call `/retry` on a failed job ID rather than submitting a duplicate job
- Check `/health` before relying on an endpoint idle for a week, since max workers drop to 0

## What the review panel asked for

- document how to lower the default cap
- Publish rate limits and 429 behaviour
- Document serverless error responses

## 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.