ModelSpec

Which models fit this job? Call decide early. Returns answer, ranked results, may_qualify, reading and ⟨275 unchanged words⟩ . Costs 1 or 2 credits a call; a credit costs $0.0053–$0.02 dependingper oncredit theby plan or pack (https://modelspec.dev/pricing/). Not a router: call it once per job or role, not per request, thenrequest; route among the modelsits itchosen recommends.models. Proxies POST https://api.modelspec.dev/v1/decide. Call decide early with a template-based Spec; refine from reading and recovery hints. Put Musts in where: these gates exclude. Put Prefers in optimize.weights: weights rank and never exclude. Unknown values go to may_qualify. Pin snapshot for reproducibility. MCP decide returns a bounded answer of at  ⟨26 unchanged words⟩  unset or "none": explain=none, limit=10, fields model_rankfields=model_rank, cost_per_task, estimates, p_best. summary or full alsosummary/full sendsadds contributions and evidence. omitted counts removed records; explain=none does not computeskips eliminations. explicitExplicit fields, including null, is sent as given. Over 16pass KBthrough. theOversize bodyresponses isuse a trimmed notice and explanation.fetch says how towith resend.explanation.fetch. decision_id is a citation, not a stored lookup. task free text is rejected; where is an array. Call shape, examplesExamples, recovery and budgets: https://modelspec.dev/agents.md. Null means not researchedunresearched or not publishedunpublished, never a guess. Every sourced value on aSourced card carriesvalues include a URL and a read date. This tool returns the origin's JSON plusReturns the origin URL;JSON itand doesURL notwithout inventinvented fields. Example. Answer: status partial, answer.kind tied ⟨26 unchanged words⟩ 

Was a ModelSpec answer reliable? After you act on an answer, say whether it was reliable, unreliable, trustworthy, untrustworthy orreliable? confusing,Returns byrecorded/not_recorded proxyingfor POSTa https://api.modelspec.dev/v1/feedback.rating Includeafter theacting; answer'sinclude decision_id. Free, no key, and no Authorization header is forwarded. Not for questions or support. Never put aNo promptprompts, a keykeys or personal details in note. The response saysnote; statusno "recorded"questions or "not_recorded"support. (storageStorage isstays off until covered by the privacy statementstatement. coversProxies it).POST https://api.modelspec.dev/v1/feedback.

Which rankinglegacy profiles exist, and what is the ranking policy?rules exist? Returns https://modelspec.dev/api/rank/profiles.json: the profiles, thewith evidence floors and the neutrality commitment. Needs an API key. Not a decision: useUse decide for that.decisions.

What does ModelSpec's card sayone aboutmodel's onecard model?say? Reads https://modelspec.dev/api/models/<model_id>.json (model_id is provider/slug). With//modelspec.dev/api/models/<model_id>.json. dataSplit splitting,mode returns only the model'sits display name from the Worker vocabulary.name. Needs an API key. Not current evidence and notor a ranking:ranking; use decide for that.decide.

Which models, on whichand platforms, does my policy allow? Checks a policy document (licence,Returns originpass, data handling) perfail modelor andundetermined per platform.model/platform Returnsfor passlicence, fail or undeterminedorigin forand eachdata row;handling; undetermined is not a pass. Needs an API key (Authorization: Bearer). Costs 5 credits a call; a credit costs $0.0053–$0.02 dependingper oncredit theby plan or pack (https://modelspec.dev/pricing/). It doesn't choose a model:model; use decide to choose among the ones thatpassing pass.rows. Proxies POST https://api.modelspec.dev/v1/policy-check with this tool'sthese arguments as the JSON body; Authorization isand forwarded.Authorization.

Legacy v1; use decide. Deprecated: the retired fixed-benchmark ranking, whoseranking; scores lag the catalogue;catalogue and newer models aremay oftenbe unranked. Which models rank highest forlead onea fixed use-case profile? Returns a shortlist with evidence_basis, and input provenance, not a quality verdict. Needs an API key (Authorization: Bearer). Costs 1 credit a call; a credit costs $0.0053–$0.02 dependingper oncredit theby plan or pack (https://modelspec.dev/pricing/). Don't call it as a fallback when decide nameshas no single leader:leader; report decide'sits tie. Present `summary_for_user` to the user unchanged and ⟨11 unchanged words⟩ of a pick. Proxies POST https://api.modelspec.dev/v1/rank with this tool's arguments as the JSONthese body.arguments.

What can a decide specSpec say? Returns the decision vocabulary: valid facet ids, benchmarks, domains, providers and task types. Compact by default, with 20 rows per page and no counts. Split-mode full details exclude per-model facts and counts. Needs an API key. Use it for missing ids, not to compare models. One vocabulary lookup per unknown facet, then decide. ReadUse `next`decide beforeto anothercompare call.models. section=starterStarter returnsincludes a ready-to-send minimal Spec and a next hint. SearchSpec. coversCompact everypages: section'sat ids,most labels,20 definitionsrows, values and synonyms.no Allcounts; tokensempty rankpages first;end aresults. content-tokenSearch subsetranks canall match.tokens Abefore misssubsets; returnsmisses theoffer nearest ids. An explicit non-starter section scopes the search. Use id or ids forSplit-mode rowfull details, or detail=full for all display details.exclude Usemodel offsetfacts and limit to page matches across sections, including full details; an empty page ends the list.counts.