Verification register AI & Agents
Current reading
Readiness band and full integer triple
AI-assisted assembly · derived results
Readiness band
Ready
Primary summary from verified readiness
- Confidence
- 74% · fresh
- Computed at
- 2026-10-29T16:46:12.186892+00:00
- Claimed
- 75
- Reported
- 67
- Verified
- 59
- Gap
- +16
Public ambition and stated capability
Observed practitioner reporting
Independently supported evidence
Claimed leads verified
Evidence strength Critical
Decision
What the current evidence supports
Human editorial judgment · 2026-10-29
Proceed with caution
- Why
- The model is live on the API with flat 1M-context pricing ($2/$10 per M tokens) and Artificial Analysis data puts it 4th overall (55), +11 points over K3, so capability evidence is real. But the open-weight claim is unfulfilled until the Nov 4 release, the license is unknown, key numbers are vendor-run, and it lags frontier closed models on pure reasoning and computer use. Provenance concerns (unsubstantiated US government allegations about K3) add procurement risk.
- Next
- Re-check after 2026-11-04 for actual weight publication, license text and independent replication of Terminal-Bench 4.0 / SWE-Marathon; pilot via API on an agentic coding workload before committing.
Constraints
Blockers
No named blocker is present in the current public projection.
Evidence summary
Derived counts
AI-assisted assembly
- Total
- 8
- Tier 1
- 2
- Tier 2
- 6
- Tier 3
- 0
- Supports
- 3
- Contradicts
- 3
- Context
- 2
- Latest observed
- 2026-10-26
Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.
Publication record
Revisions
Initial public reading
- 2026-10-29 Initial recorded band: ready.