We're inviting Anthropic to put the real Mythos 5 on our open benchmark
Verified live by HTTP probe across 22 observations. Pulse has been flat over the last 30 days. Observation is liveness-only: no tracked press, repository, or traffic rank — existence is confirmed, momentum is unobserved.
HTTP liveness
realSignal telemetry
4 families · 1 observed · observed onlyrealvuln.comedge
direct originserver
AmazonS3Tranco rank
unranked
Analyst read
generated from the observed metrics above · no external narrativeWe're inviting Anthropic to put the real Mythos 5 on our open benchmark's listed site resolved live when we last probed it on Aug 24, 2026, so we mark it verified alive. It has fewer than two observed signal families, so no Pulse Index is composed — the evidence strip shows what is and isn’t present. Observed footprint: 5 functional tags. It has no curated competitive-alternative edges yet, so the comparable set below is functional tag-similarity — read it with the confidence flags.
Comparable set
curated alternatives → functional tag-similarity| Company | Basis | Overlap | Pulse | 30d Δ | Status |
|---|---|---|---|---|---|
| AI Security Leaderboardai-tools | similar | 100% | 40.0 | — | ● live |
| Agent Arenaai-tools | similar | 83% | 40.7 | 0.0 | ● live |
| pumaDBai-tools | similar | 83% | 40.0 | 0.0 | ● live |
| Agent Memory Leaderboardai-tools | similar | 80% | 32.8 | — | ● live |
| A Terrible Way to Consume Hacker Newsai-tools | similar | 80% | 40.0 | 0.0 | ● live |
| TokenPathai-tools | similar | 80% | 40.0 | +9.6 | ● live |
| bleeding edgeai-tools | similar | 80% | 40.0 | 0.0 | ● live |
| FirewallAuthoritysecurity | tag-overlap | 80% | 40.0 | 0.0 | ● live |
Confidence reflects source and category coherence: curated = an editorially linked alternative; similar = strong functional tag-overlap in the same category; tag-overlap = shared tags only and may be noisy. Overlap is the shared functional-tag ratio (Jaccard) for tag-matched rows. Pulse is the observed-activity index real and 30d Δ its signed month change; a dash means insufficient observed history, never zero.
Structure
observed attributesFunding & team
not in sourceNo disclosed funding in our sources. Funding data covers only ~1% of the corpus (462 rounds across 79K products), so absence here is unknown, not zero— we don't estimate a number we can't source.