Sign in
Pending benchmarksTracking since Nov 9, 2026

Is OrcaC2 1.0 state of the art?Pending benchmarks.Announced Nov 7, 2026; no independent results yet.

OrcaCyber's incident-decision model on OrcaRouter: a closed language-vision systemone model that takes a text/image state plus typed questions and returns calibrated machine-readable verdicts (probability and confidence) in one forward pass for triage and containment workflows (Act / Abstain / Escalate). Served at POST /v1/systemone as orca/orcac2-1.0. No open weights are published.

Announced facts

The details so far

Lab
OrcaRouter
Announced
Nov 7, 2026
API
Live · orca/orcac2-1.0
Open weights
Closed; only Nano in the OrcaCyber line is listed with open weights
Lab / product line
OrcaCyber Decide model, served through OrcaRouter (same catalog lab as OrcaCyber Nano 1.0)
Role
Incident decisioning for MDR/IR-style triage and containment; structured verdicts rather than free-form chat
Interface
Typed systemone API: observation state plus one or more typed questions; parallel answers; optional reasoning / reasoning_effort
Modalities
Text and image inputs; text structured outputs (OrcaRouter card)
Access
Public OrcaRouter API with an API key; model id orca/orcac2-1.0. Official site lists Preview / Copy ID for OrcaC2, not the verified-access or early-access badges used for Zero/Ultra. Offensive-capability verification language on orcacyber.ai applies to offensive models, not this Decide model
Pricing (lab)
Per-call tiers on OrcaRouter: $0.0010 up to 16K input tokens, $0.0100 above that, $0.0300 when an image is included
Latency (lab)
OrcaRouter reports p50 TTFT about 346 ms over a recent window
Context
Not stated on the OrcaC2 OrcaRouter card (context_window listed as 0). orcacyber.ai advertises up to 1M tokens as the OC-01 family maximum — not treated as a verified OrcaC2-specific figure here
Related
Keep separate from OrcaCyber Nano 1.0 and from other orcacyber-* OrcaRouter IDs (Zero, Ultra, Flash, etc.)
Date evidence
OrcaRouter public model release_date 2026-11-07; live traffic and latency series on that platform. @OrcaCyberHQ post on Nov 9 is later. An older third-party '2026-10-30' byline is not used

Tags marked unsure are not stated on the lab's own pages.

Next steps

What would let us call it

Verdicts use only independent boards we already track. Two consecutive snapshots are required before any crown.

  • Publicly announced with a source
  • Usable by the public (API or weights)
  • Listed on an independent board we track
  • Enough results for a rank in a category
  • Holds across two snapshots (crowns only)
Self-reported

What the lab claims

Self-reported. Shown for context; not used in any verdict.

Lab claim: returns calibrated machine-readable verdicts with a probability and a confidence in a single forward pass, without chain-of-thought generation (OrcaRouter model card).

Lab claim: built for MDR and IR triage/containment decisions framed as Act / Abstain / Escalate (orcacyber.ai).

Receipts

Sources

LABOrcaCyber site: OrcaC2 Decide modelorcacyber.ai/LABOrcaRouter: OrcaC2 1.0 model pageorcarouter.ai/models/orca/orcac2-1.0 · Nov 7, 2026LABOrcaCyberHQ announcement on Xx.com/OrcaCyberHQ/status/2119821023927521634 · Nov 9, 2026

Tracking continues until an independent board lists it.