{
  "listings": [
    {
      "slug": "celeris-1-decision",
      "name": "Celeris-1 Decision",
      "company": "Marqo Inc",
      "grade": "D",
      "fit": "yes",
      "why": "score returns a probability per level, the probability-weighted average level and a legend, up to 64 levels. Thresholding guidance is generic.",
      "source": "https://docs.celeris.ai/decisions.md",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/celeris-1-decision"
    },
    {
      "slug": "cloudflare-clef",
      "name": "Clef",
      "company": "Cloudflare, Inc.",
      "grade": "B",
      "fit": "yes",
      "why": "score questions over ordered options return a probability per option. No threshold guidance from Cloudflare.",
      "source": "https://huggingface.co/Cloudflare/clef",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/cloudflare-clef"
    },
    {
      "slug": "decider",
      "name": "Decider",
      "company": "Mark Marosi (Mapika)",
      "grade": "BB",
      "fit": "yes",
      "why": "score over 2 to 10 described levels returns a probability per level and the expected level, with a fitted temperature per answer type.",
      "source": "https://github.com/Mapika/decider#readme",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/decider"
    },
    {
      "slug": "decision-3",
      "name": "Decision 3.0",
      "company": "vLLM Semantic Router project",
      "grade": "B",
      "fit": "yes",
      "why": "score questions return a probability per level; the readout temperature is 1.0 with no calibration figure.",
      "source": "https://huggingface.co/vllm-sr/d3",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/decision-3"
    },
    {
      "slug": "elephant-1",
      "name": "elephant-1",
      "company": "Baseten Labs, Inc.",
      "grade": "D",
      "fit": "yes",
      "why": "score over 2 to 10 ordered levels returns a probability per level and the expected level; a Brier term in training discourages overconfidence.",
      "source": "https://huggingface.co/baselabs/elephant-1",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/elephant-1"
    },
    {
      "slug": "joey-1",
      "name": "joey-1",
      "company": "Baseten Labs, Inc.",
      "grade": "C",
      "fit": "yes",
      "why": "score over 2 to 10 ordered levels returns a probability per level and the expected level; a Brier term in training discourages overconfidence.",
      "source": "https://huggingface.co/baselabs/joey-1",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/joey-1"
    },
    {
      "slug": "liquid-d1",
      "name": "Liquid d1",
      "company": "Liquid AI, Inc.",
      "grade": "C",
      "fit": "yes",
      "why": "score returns the probability-weighted position and a probability per level; the guide's scoring and reranking examples set thresholds on the probabilities.",
      "source": "https://docs.liquid.ai/lfm/models/d1",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/liquid-d1"
    },
    {
      "slug": "mammoth-1",
      "name": "mammoth-1",
      "company": "Baseten",
      "grade": "D",
      "fit": "yes",
      "why": "score over 2 to 10 ordered levels returns a probability per level and the expected level; a Brier term in training discourages overconfidence.",
      "source": "https://huggingface.co/baselabs/mammoth-1",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/mammoth-1"
    },
    {
      "slug": "microsoft-decision-1",
      "name": "Microsoft-Decision-1",
      "company": "Microsoft",
      "grade": "E",
      "fit": "yes",
      "why": "score over ordered levels returns an expected score and probabilities; Learn says to validate on your own labelled examples and prefer scores for ordering and thresholds over absolute ratings.",
      "source": "https://learn.microsoft.com/en-us/azure/foundry/foundry-models/how-to/use-foundry-models-microsoft-decision",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/microsoft-decision-1"
    },
    {
      "slug": "nace-drex",
      "name": "Drex 1.5",
      "company": "Nace AI, Inc.",
      "grade": "BB",
      "fit": "yes",
      "why": "score returns the expected score, a legend, a probability per level and confidence; levels are the criteria array in order.",
      "source": "https://console.nace.ai/docs/drex/guides/questions",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/nace-drex"
    },
    {
      "slug": "openai-decisions-api",
      "name": "OpenAI Decisions API",
      "company": "OpenAI OpCo, LLC",
      "grade": "A",
      "fit": "yes",
      "why": "score over 2 to 10 ordered levels returns the probability-weighted average level, probabilities and confidence.",
      "source": "https://developers.openai.com/api/docs/guides/decisions",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/openai-decisions-api"
    },
    {
      "slug": "pplx-decider",
      "name": "pplx-decider",
      "company": "Perplexity AI, Inc.",
      "grade": "BB",
      "fit": "yes",
      "why": "score over up to 10 levels returns the probability-weighted score, legend and probabilities; confidence formula is documented.",
      "source": "https://docs.perplexity.ai/docs/decisions/quickstart",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/pplx-decider"
    },
    {
      "slug": "pplx-decider-flash",
      "name": "pplx-decider-flash",
      "company": "Perplexity AI, Inc.",
      "grade": "B",
      "fit": "yes",
      "why": "score over up to 10 levels with expected score, legend and probabilities; same documented confidence formula.",
      "source": "https://docs.perplexity.ai/docs/decisions/quickstart",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/pplx-decider-flash"
    },
    {
      "slug": "scaledown",
      "name": "ScaleDown",
      "company": "Carbon ScaleDown, Inc.",
      "grade": "E",
      "fit": "yes",
      "why": "score over 2 to 10 levels returns the probability-weighted expected value and a probability per level, with the calculation worked through in the docs.",
      "source": "https://docs.scaledown.ai/api-reference/decisions",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/scaledown"
    },
    {
      "slug": "sprag-spev",
      "name": "SPEV",
      "company": "Sprag AI, Inc.",
      "grade": "D",
      "fit": "yes",
      "why": "score over 2 to 10 levels returns the expected level, probabilities and confidence; vendor figures only.",
      "source": "https://sprag.ai/docs/decisions/api",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/sprag-spev"
    },
    {
      "slug": "sqwish-kettei",
      "name": "kettei",
      "company": "Sqwish Ltd, trading as Sqwish Labs",
      "grade": "C",
      "fit": "yes",
      "why": "ordinal decisions over 2 to 255 levels with a rubric line each return probabilities and expected_level.",
      "source": "https://docs.sqwish.ai/llms-full.txt",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/sqwish-kettei"
    },
    {
      "slug": "typesafe-jev",
      "name": "Jev",
      "company": "TypeSafe AI, Inc.",
      "grade": "B",
      "fit": "yes",
      "why": "score with probabilities per level, expected score and confidence, plus confidence-routing guidance for answers split across levels; fraud triage on 100 transactions and routing on 80 prompts ran against Strands with Jev stronger on fraud.",
      "source": "https://docs.typesafe.ai/confidence",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/typesafe-jev"
    },
    {
      "slug": "convai-laya",
      "name": "Laya",
      "company": "Convai Innovations Pvt. Ltd.",
      "grade": "BB",
      "fit": "caveat",
      "why": "score questions are supported in one pass at 33 ms, with answer_confidence to gate on, but issue 131 reports laya-multilingual never selecting the first-listed score option and calibration error is high.",
      "source": "https://huggingface.co/convaiinnovations/laya",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/convai-laya"
    },
    {
      "slug": "drex-max-v1",
      "name": "Drex Max v1",
      "company": "Nace AI, Inc.",
      "grade": "C",
      "fit": "caveat",
      "why": "score over ordered levels with expected score and probabilities. Vendor figures only; licence cap.",
      "source": "https://huggingface.co/nace-ai/drex-max-v1",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/drex-max-v1"
    },
    {
      "slug": "granite-switch",
      "name": "Granite Switch 4.1",
      "company": "International Business Machines Corporation",
      "grade": "B",
      "fit": "caveat",
      "why": "Ordered labels can be scored, but there is no score type, legend or expected level; the caller derives them from label logprobs.",
      "source": "https://huggingface.co/ibm-granite/granite-switch-4.1-8b",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/granite-switch"
    },
    {
      "slug": "jaredpalmer-kev",
      "name": "Kev",
      "company": "Jared Palmer",
      "grade": "BB",
      "fit": "caveat",
      "why": "score over 1 to 255 levels with expected level and probabilities; issue 161 measured a slot-0 logit offset on identical score options, fixed in PR 236, so pin v1.0 or later weights.",
      "source": "https://github.com/jaredpalmer/kev#readme",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/jaredpalmer-kev"
    },
    {
      "slug": "jev-27b-vl",
      "name": "JEV-27B-VL",
      "company": "AutoTrust AI Lab",
      "grade": "C",
      "fit": "caveat",
      "why": "Rate 0 to 5 on text and images with a probability per level; vendor figures only.",
      "source": "https://huggingface.co/autotrust/JEV-27B-VL",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/jev-27b-vl"
    },
    {
      "slug": "rightnow-system-zero",
      "name": "RightNow System Zero",
      "company": "RightNow AI, Inc.",
      "grade": "B",
      "fit": "caveat",
      "why": "score returns an expected level, legend and probabilities, flagged uncalibrated.",
      "source": "https://www.rightnowai.co/llms-full.txt",
      "confidence": "medium",
      "url": "https://www.anchorterminal.com/tools/rightnow-system-zero"
    },
    {
      "slug": "strands-decider",
      "name": "Strands Decider 2B",
      "company": "Amazon Web Services, Inc.",
      "grade": "B",
      "fit": "caveat",
      "why": "score with expected level and probabilities; fraud triage scored 61 of 100 matching words in option descriptions, and the card says confidence bands hold for short classification only.",
      "source": "https://builder.aws.com/content/3K787ZfsalsR9818hYPXpScY3J6/system-one-models-strands-decider-b-on-the-same-tests-i-ran-on-jev-last-week",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/strands-decider"
    },
    {
      "slug": "vela",
      "name": "Vela 2.0",
      "company": "vLLM Semantic Router project (with KR Labs)",
      "grade": "BB",
      "fit": "caveat",
      "why": "score over 2 to 10 levels returns the expected level and distribution; USAGE.md says Choice and Score confidence are distribution statistics, not calibrated probabilities.",
      "source": "https://huggingface.co/vllm-sr/Vela-2.0-4B/blob/main/USAGE.md",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/vela"
    },
    {
      "slug": "gliclass",
      "name": "GLiClass",
      "company": "Knowledgator Engineering Ltd.",
      "grade": "C",
      "fit": "no",
      "why": "No ordered score type or expected level; labels are unordered and scored independently.",
      "source": "https://docs.knowledgator.com/docs/frameworks/gliclass/usage/",
      "confidence": "high",
      "url": "https://www.anchorterminal.com/tools/gliclass"
    }
  ],
  "marks": "yes: the docs show it does the job as the success test says. caveat: it does the job with a limit an agent should know. no: it doesn't do this job. unchecked: we couldn't read it.",
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "attributionRequired": true,
    "credit": "CC BY 4.0, credit Anchor Terminal and link the page or API address you took it from (https://www.anchorterminal.com/terms/#data)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-11-16",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "0.8",
    "openapi": "https://www.anchorterminal.com/openapi.json",
    "preview": false,
    "run": "2026-10-01",
    "runLabel": "October 2026 research run",
    "terms": "https://www.anchorterminal.com/terms/#data"
  },
  "useCase": {
    "accounts": 4,
    "after": [
      "support.set-priority",
      "human-review"
    ],
    "area": "models",
    "asks": [
      "how urgent is this",
      "what fraud risk level is this transaction",
      "rate this on a 1 to 5 rubric",
      "how severe is this incident"
    ],
    "before": [
      "support.fetch-ticket",
      "payments.fetch-transaction"
    ],
    "blockers": [
      "expected score rounds a critical item into a lower queue when probability is spread",
      "a one-point rubric cutoff hides borderline cases"
    ],
    "capabilities": [
      "inference.decision"
    ],
    "category": "decision-models",
    "categoryName": "Decision models",
    "checked": "2026-11-12",
    "companies": {
      "fit": 15,
      "fitWithLimit": 7,
      "tagged": 30
    },
    "competition": "crowded",
    "fits": [
      "celeris-1-decision",
      "cloudflare-clef",
      "decider",
      "decision-3",
      "elephant-1",
      "joey-1",
      "liquid-d1",
      "mammoth-1",
      "microsoft-decision-1",
      "nace-drex",
      "openai-decisions-api",
      "pplx-decider",
      "pplx-decider-flash",
      "scaledown",
      "sprag-spev",
      "sqwish-kettei",
      "typesafe-jev"
    ],
    "fitsWithLimit": [
      "convai-laya",
      "drex-max-v1",
      "granite-switch",
      "jaredpalmer-kev",
      "jev-27b-vl",
      "rightnow-system-zero",
      "strands-decider",
      "vela"
    ],
    "id": "decision-models.score-rubric",
    "listings": {
      "caveat": 8,
      "no": 1,
      "notRead": 1,
      "tagged": 33,
      "unchecked": 0,
      "yes": 17
    },
    "name": "Place an item on an ordered scale",
    "status": "draft",
    "success": "A probability per level and an expected score. Measured: the chosen reading places at least 85 percent of the caller's labelled items in the right level and no critical item below the urgent cut.",
    "url": "https://www.anchorterminal.com/api/v1/use-cases/decision-models.score-rubric.json",
    "weight": 8,
    "weightBasis": "4 of 41 accounts; the AWS fraud and delivery tests, the Kev score-order issue and the Dataiku score threshold finding"
  }
}
