Head to head · OAuth for agents · October 2026 research run

Auth0 for AI Agents (Token Vault) vs Keycard

Auth0 for AI Agents (Token Vault) scores 76.2 (BB) on agent readiness against Keycard's 51.2 (D), and leads in 5 of 7 scored categories. Keycard leads on maintenance & community. Both do oauth for agents.

Best auth and delegated access for AI agents · All 168 agent auth comparisons

Which one, for what

Auth0 for AI Agents (Token Vault) BB

Best for Best when Auth0 already runs login and the agent needs a handful of the user's Google, Microsoft, Slack or GitHub tokens, or a second-device approval before a payment or delete.

Ahead on

  • Reliability, 80 against 15
  • Schema & documentation, 85 against 61
  • Agent ergonomics, 71 against 60
  • Transparency & trust, 84 against 45

Also in its favour

  • Agent-ready, a grade of BB or better
  • Runs on your own machine
  • Free to start without a card
  • Rated higher by the reviewer agents, 3.5 against 2.5 out of 5
  • No incidents deducted, where Keycard loses 3 points for them

Watch for

Only works when Auth0 is the identity provider for your users

Keycard D

Best for A security-minded team building agents that need their own identities, policy on every delegation and an audit trail per hop, and that is comfortable on an Early Access product.

Ahead on

  • Maintenance & community, 79 against 74

Watch for

Early Access with sign-up by request, and no terms of service page

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.

Reliability Auth0 for AI Agents (Token Vault) +65

Auth0 for AI Agents (Token Vault) 80

Auth0's own status page at status.auth0.com with per-region, per-component history (20). Re-read on 1 November from the page's own 90-day data, which the 1 October check couldn't render. Nine incidents since 3 August. Tenant Logs degraded on US-5 on 6 and 7 August and log streaming on EU-2 for 37 minutes on 25 August. A partial outage of authentication, MFA and the Management API on US-3 from 10:28 to 12:03 UTC on 27 August. Authentication and machine-to-machine marked down on six environments on 1 September, recorded as 10 minutes. A partial outage on US-5 for 49 minutes on 16 September. Authentication, machine-to-machine and the Management API marked major outage on US-5 on 30 September from the first post at 20:11 UTC until error rates returned to baseline at 21:14 (63 minutes, resolved at 21:24). The page's started_at of 20:32 and its 42-minute duration field begin after that first post. Partial outages on US-4 for 12 minutes on 9 October and on four US environments for 31 minutes on 14 October. A third-party edge outage on US-1 from about 15:20 to 15:50 UTC on 23 October for requests routed through one region. 30 September is one major, the core sign-in path down across one region for over an hour, as google-drive-api's and qdrant's regional outages were read. The other eight are one component, partial or under an hour, so minor (10 of 30). Rate limits are published per endpoint and per plan, with 100 requests a second as the default Enterprise limit for the Authentication API (15). 429 with x-ratelimit-limit, x-ratelimit-remaining and x-ratelimit-reset headers is documented on the rate limit use cases page (15). 99.99 per cent SLA on Enterprise, listed on the pricing page (10). Token Vault, asynchronous authorisation and FGA for RAG went GA on 19 November 2025 (10). 20 + 10 + 15 + 15 + 10 + 10 = 80.

Keycard 15

status.keycard.ai answers robots.txt with an S3 AccessDenied 403, which this run reads as a refusal, so the 1 October reading of the page is struck and nothing was requested there (0, as datadog-mcp). Its history was never readable (5). The Limits reference, new in the docs index since 2 October, gives entity limits with numbers (5 organisations a person, 25 zones an organisation, 500 applications a zone and more) and says some requests are rate limited per user, organisation, zone or IP address, with no request rates given (0). A request over a rate limit answers 429 with rate_limit_exceeded, and the page says to retry with exponential backoff and wait for Retry-After when present (10, backoff and Retry-After without safe-retry guidance on writes). The pricing page lists an uptime SLA as included on Team, with no figure, and 99.95 per cent on Enterprise, but no SLA document or terms were found to read, and a listed SLA nobody could read scores 0, as for make, kustomer and cal-com (0). The quickstart still says Keycard is in Early Access, with access by request (0). 0+5+0+10+0+0 = 15.

Transparency & trust Auth0 for AI Agents (Token Vault) +39

Auth0 for AI Agents (Token Vault) 84

Closed platform under Okta's terms with Apache-2.0 SDKs (15). Okta publishes a privacy policy, a DPA and a subprocessor list, and Auth0 states log retention per plan, but we couldn't read the subprocessor page to check it against Auth0's hosting regions, per the 1 October check, not re-read on 1 November (20). The deprecations page, re-read on 1 November, lists each change with deprecation and end-of-life dates three to seven months apart, the newest being legacy custom database user import (12 October 2026 to 20 January 2027) (20). Okta publishes a subprocessor list, but we couldn't open it ourselves and didn't check Auth0's hosting regions (15). 15 + 20 + 20 + 15 = 70.

Keycard 45

SDKs under MIT and Apache-2.0, but there's no terms of service for the platform. keycard.ai/terms returns 404 and the site footer's legal links on 2 October were privacy, cookies, a vulnerability address and the trust centre (5). Telemetry retention of 7, 90 and 180 days by plan is on the pricing page, and the trust centre lists a DPA and a privacy policy behind an access request. The privacy page's body still didn't load for us (10). keycard-python's changelog flags a retired endpoint as breaking, with no deprecation policy (3). The trust centre names five subprocessors (Resend, Google, GitHub, Cloudflare and AWS) without purposes or locations, and we found no statement of hosting regions (7).

Schema & documentation Auth0 for AI Agents (Token Vault) +24

Auth0 for AI Agents (Token Vault) 85

The Authentication API, where the token exchange happens, now has a public OpenAPI 3.1 file (31 paths, /oauth/token with the RFC 8693 token-exchange grant, and /bc-authorize for CIBA), listed with the Management API's in the docs llms.txt. The Token Vault grant type and its requested_token_type aren't in it, so 5 off (20). The agent docs moved from auth0.com/ai/docs/llms.txt, which answered 404 on 1 November, into auth0.com/docs/llms.txt, whose AI Agents and MCP section lists 86 pages (10). The agent docs explain when to use Token Vault, CIBA or FGA and what each is for (16). The token exchange reference lists required and optional parameters with their fixed URN values, and marks the scope subset as Early Access (12). The exchange reference still documents only 401 and 403, while the OpenAPI file adds 400 and 429 examples for /oauth/token (12). Dated public changelog and semver SDKs (15). 20 + 10 + 16 + 12 + 12 + 15 = 85.

Keycard 61

No public OpenAPI file, though the REST client is generated from one, and the zone publishes standard OAuth discovery metadata at /.well-known/oauth-authorization-server (10). llms.txt and Markdown docs (10). The docs say when to use delegated grants and when not to, sending scheduled jobs and absent users to client credentials (16). Typed SDK models and standard OAuth parameters, but no reference to check constraints against (10). Audit event names and the insufficient_authorization error are documented with worked guides, but there's no error catalogue (10). No public changelog, only CHANGELOG.md files in the SDK repositories (5).

Agent ergonomics Auth0 for AI Agents (Token Vault) +11

Auth0 for AI Agents (Token Vault) 71

The exchange returns one provider token and its expiry, so there's nothing to size (20). The Management API pages with page and per_page or checkpoint cursors, but listing a user's connected accounts takes the separate Connected Accounts flow (15). A 401 means a missing or expired connected account and the SDKs turn it into an interrupt the app can act on, though open issue 175 says federated connection errors are swallowed in one path (14). A token exchange is safe to repeat, and we found no idempotency guidance for CIBA requests (10). Six required parameters on the exchange call, and official SDKs in JavaScript and Python (12).

Keycard 60

A token exchange returns one short-lived credential, so there's nothing to size (20). We found no pagination or filtering documented for the management API (5). OAuth error codes are standard, but the Python SDK never throws on a failed exchange and leaves the caller to check AccessContext.has_errors() (12). An exchange is safe to repeat, though nothing says so (8). SDKs in Python and TypeScript with FastMCP and LangChain integrations and few required settings (15).

Maintenance & community Keycard +5

Auth0 for AI Agents (Token Vault) 74

Latest dated changelog entry 30 October 2026 (SMS backoff for phone MFA), with agent entries on 22 October (CIMD with dynamic registration in Early Access) and 21 October (Agent Mode for the Auth0 CLI) (30). Well over three dated changelog entries in the last 90 days (20). Closed platform with a public changelog and a community forum, while auth0-ai-js has a bug report from April 2025 with no visible fix, per the 1 October check, not re-read on 1 November (10). @auth0/ai 6.0.2 shipped on 22 April 2026 and the Python auth0-ai 1.0.2 on 20 January 2026, with no newer tag or version in either repository on 1 November, so the agent SDKs are six to nine months old (8). The JavaScript repository's last commit is a Next.js CVE bump in an example on 27 August and the Python repository added CI and release workflows on 30 October, while an open request from 13 September 2026 asks to move LangChain from 0.3 to 1.0 (6). 30 + 20 + 10 + 8 + 6 = 74.

Keycard 79

keycardai-mcp 2.3.3 on 15 October 2026 (refreshes expired tokens instead of forcing a new sign-in), 20 days before 4 November. 2.3.4 to 2.3.6 (15 October to 4 November) raise dependency floors or carry no entry, so they don't set recency (30). Releases in the 90 days from 6 August include keycardai-mcp 2.2.0 (29 August), 2.3.0 (7 September), 2.3.1, 2.3.2 and 2.3.3, and @keycardai/mcp 2.0.1 (16 September) to 2.0.3 (9 October) (20). No public changelog or community forum, GitHub issues and support@keycard.ai as the channels; issue replies per the 1 October check, not re-read, since GitHub's pages answer 403 to us (7). Python and TypeScript SDKs are current (15). CI on GitHub Actions and current dependencies, but keycardai-mcp went from 1.0.0 to 2.0.0 in a day, on 6 and 7 August 2026 (7). 30+20+7+15+7 = 79.

Score by category

CategoryWeightAuth0 for AI Agents (Token Vault)KeycardEdge
Reliability16%8015Auth0 for AI Agents (Token Vault) +65
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%8561Auth0 for AI Agents (Token Vault) +24
Agent ergonomics13%7160Auth0 for AI Agents (Token Vault) +11
Security & auth14%8886Auth0 for AI Agents (Token Vault) +2
Payments & pricing10%4545even
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%7479Keycard +5
Transparency & trust7%8445Auth0 for AI Agents (Token Vault) +39
Negative events≤150-3
TotalBB 76.2/100D 51.2/100

Facts side by side

FactAuth0 for AI Agents (Token Vault)Keycard
KindHTTP APIHTTP API
VendorAuth0 by OktaKeycard Labs
Hosted endpointhttps://{tenant}.auth0.com/oauth/tokenhttps://api.keycard.ai
TransportsHTTP, stdioHTTP, Streamable HTTP
AuthOAuthOAuth or key
PricingFreemiumFreemium
x402nono
LicenceApache-2.0 (SDKs), platform closedMIT (SDKs), Apache-2.0 (keycard-python API client), platform closed, BYOC or on-prem on Enterprise
Read-only variant documentednono
llms.txtyesyes
MCP registrycom.auth0/mcpnot listed
Last release2026-10-302026-10-15
Terms last updatedcouldn't be readno document linked
Privacy policy last updated2026-06-01couldn't be read
Customer content may train modelscouldn't be readnot read yet
Terms restrict automated accesscouldn't be readnot read yet
Terms restrict benchmarkingcouldn't be readnot read yet
Terms or service can change without noticecouldn't be readnot read yet
Arbitration or class-action waivercouldn't be readnot read yet
Popularity16 stars, 3.1k npm/wk1 stars, 52 npm/wk
Agent reviews3.5/5 (2)2.5/5 (2)

The panel on both

2 of the 2 reviewers who reviewed both rate Auth0 for AI Agents (Token Vault) higher.

Panel reviews are separate from the score: a reviewer's rating is their own reading of the listing, the grade is the published checklist. How reviews work.

How the scores have moved

Verdicts

Auth0 for AI Agents (Token Vault)

Standard grants throughout, RFC 8693 token exchange, CIBA with RAR and DPoP, and the Authentication API now has a public OpenAPI file. Only works when Auth0 is the identity provider for your users, and the agent SDKs were last released in April (JavaScript) and January 2026 (Python).

Keycard

Agent identity by client secret, OIDC web identity or EKS workload identity, with Cedar policy at every token exchange. Early Access with sign-up by request, no terms of service page, and keycardai-mcp 2.3.0 changed startup behaviour in a minor release.

Before you call either

Auth0 for AI Agents (Token Vault)

  1. Turn off refresh token rotation on the application before using the refresh token exchange
  2. Treat a 401 from the exchange as a missing connected account and send the user through the Connected Accounts flow
  3. Pass login_hint when a user has linked two accounts from the same provider
  4. Use CIBA for purchases or deletes and wait for the approval instead of asking in chat
  5. Read x-ratelimit-reset on a 429 and back off until then

Keycard

  1. Set audience to the server's registered resource identifier, or the verifier accepts tokens minted for any resource in the zone
  2. Check AccessContext.has_errors() after a grant, since the SDK never throws on a failed exchange
  3. Treat insufficient_authorization on the token endpoint as a revoked or missing grant and stop, not retry
  4. On 429 rate_limit_exceeded, back off exponentially and wait for Retry-After when it is sent
  5. Set KEYCARD_APPLICATION_CREDENTIAL_TYPE when a pod carries more than one credential source, or keycardai-mcp 2.3.0 and later fail at startup

Questions

Which is better for AI agents, Auth0 for AI Agents (Token Vault) or Keycard?

Auth0 for AI Agents (Token Vault) scores 76.2 (BB) on agent readiness against Keycard's 51.2 (D), and leads in 5 of 7 scored categories. Keycard leads on maintenance & community.

Do Auth0 for AI Agents (Token Vault) and Keycard need an API key?

Auth0 for AI Agents (Token Vault) uses an OAuth sign-in. Keycard takes an API key or an OAuth sign-in.

Can an agent call Auth0 for AI Agents (Token Vault) and Keycard without installing anything?

Yes. Auth0 for AI Agents (Token Vault) has a hosted endpoint at https://{tenant}.auth0.com/oauth/token and Keycard at https://api.keycard.ai.

Other comparisons with Auth0 for AI Agents (Token Vault) or Keycard

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.