Sign in
Pending benchmarksTracking since Oct 21, 2026

Is Clara state of the art?Pending benchmarks.Announced Oct 21, 2026; no independent results yet.

Saanora's decision model. Send a record (text or JSON) and up to 64 typed questions, and it returns a probability for every allowed answer: yes or no, one of up to 255 options, or a point on a scale. It never returns free text. Available on the Saanora Platform API, billed per input token only.

Announced facts

The details so far

Lab
Saanora
Announced
Oct 21, 2026
API
Live · clara-1
Open weights
Unknown
Question types
Yes/no, choice (up to 255 options) and score; up to 64 questions per request
Context
16,384 tokens per record
Modes
fast (default, reads the record once) and consistent (re-reads it per question)
Price
$0.50 per 1M input tokens, $0 output, no per-decision fee; prepaid credit from $5
API
POST /v1/decisions; also accepts the open System One format at /v1/systemone

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: 93.55% on AgentDecisionBench tool routing (Saanora lists Jev 1.13 at 93.6%) and 83.6 on JevBench-mini coherence.

Lab claim: about 0.1 s for five decisions about one record, with 0% malformed answers.

Receipts

Sources

LABSaanora Platform models pageapi.saanora.com/models · Oct 21, 2026LABSaanora Platform changelogapi.saanora.com/changelog · Oct 21, 2026LABSaanora announcement on Xx.com/SaanoraLabs/status/2112936765531275384 · Oct 21, 2026

Tracking continues until an independent board lists it.