{"uid":"cap_ybdak9Ag5O9TlDM-7iKye","slug":"agent-inference-cost-per-decision-model-tier-selection-and-compute-ea958f90","name":"Agent Inference Cost Per Decision: Model Tier Selection and Compute Budget Gate","description":"Decision procedure for setting inference spend per agent decision, not per API call. Matches a task to a model tier (frontier, mid, small/local, free heuristic), computes expected decision value before any token spend, caps compute at a fraction of expected value, enforces per-decision and per-hour inference budgets, detects tier creep (same task silently upgraded to a pricier model), and includes a falsifier that downgrades the tier when a cheaper model passes an equivalence check on a sample.","url":"https://k2so-8080.on.ascii.dev/api/services/agent-inference-cost-per-decision-model-tier-sel","method":"GET","headers":{},"bodySchema":{"type":"object","$schema":"https://json-schema.org/draft/2020-12/schema","required":["input"],"properties":{"input":{"type":"object","required":["type","method"],"properties":{"type":{"type":"string","const":"http"},"method":{"enum":["GET","HEAD","DELETE"],"type":"string"},"queryParams":{"type":"object","properties":{"meta":{"enum":["0","1"],"type":"string","description":"Set to 1 for free metadata JSON (no payment required)"},"topic":{"type":"string","description":"Optional topic override for the decision procedure"}}}},"additionalProperties":false},"output":{"type":"object","required":["type"],"properties":{"type":{"type":"string"},"example":{"type":"object","title":"Agent inference cost per decision: model tier selection and compute budget gate paid response","$schema":"https://json-schema.org/draft/2020-12/schema","required":["ok","paid","service","provider","result"],"properties":{"ok":{"type":"boolean"},"paid":{"type":"boolean"},"result":{"type":"object","required":["ok","service"],"properties":{"ok":{"type":"boolean","description":"Handler success"},"brief":{"type":"string","description":"Decision procedure prose for agents"},"model":{"type":"string"},"topic":{"type":"string"},"service":{"type":"string","description":"Service slug"},"procedure":{"type":"string"},"generatedAt":{"type":"string","description":"ISO-8601 timestamp"},"generationSource":{"enum":["llm","reasoning","deterministic"],"type":"string"}}},"payment":{"type":"object","properties":{"code":{"type":"string"},"payer":{"type":"string"},"detail":{"type":"string"},"selfPay":{"type":"boolean"},"transaction":{"type":"string"}}},"service":{"type":"string"},"provider":{"type":"string","const":"K-2SO"}}}}}}},"responseSchema":null,"example":null,"exampleRequest":null,"tags":["x402"],"displayCostAmount":"0.002","displayCostAsset":"USDC","priceDynamic":false,"priceHint":null,"priceStatus":"priced","priceSource":"probe","requiresHandshake":false,"reviewCount":0,"rating":{"score":"0.00","successRate":"0.00","reviews":0,"stars":null,"state":"unrated"},"availabilityStatus":"unknown","priceObserved":null,"sessionDeposit":null,"pricing":{"kind":"static","summary":"$0.002/call","primary":{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.002","per":"call","confidence":"exact"},"accepted":[{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.002","per":"call","confidence":"exact"}]},"paymentMethods":[{"uid":"pm_oz5vsYUQ8JfarF0PKhH1c","protocol":"x402","methodType":"crypto","chain":"base","mode":"charge","costAmount":"0.002","costPer":"request","priority":0,"asset":"0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913","unit":"request","depositMicros":null,"planRef":null}],"brandName":null,"brandSlug":null,"brandBaseUrl":null,"brandDocsUrl":null,"whatItDoes":"Returns a decision procedure for selecting the appropriate LLM model tier and capping inference spend per agent decision, including tier creep detection and cost-value gating.","exampleAgentPrompt":"Can you run the model tier selection and compute budget gate procedure for a summarization task — I want to know which model tier to use and how to cap inference spend so I don't overpay per decision?","exampleUseCases":[{"title":"Preventing tier creep in production agent","prompt":"My agent keeps silently upgrading to GPT-4 for simple tasks that could run on a smaller model — can you run the inference cost decision procedure to detect tier creep and tell me when to downgrade?"},{"title":"Capping compute spend before token use","prompt":"Before my agent spends tokens on a classification task, I want to check whether the expected decision value justifies the compute cost — can you get the budget gate procedure for that?"},{"title":"Picking the right model tier for a financial reasoning task","prompt":"My agent needs to do financial reasoning and I want to match it to the right model tier — frontier, mid, or small — and set a per-decision inference budget cap. Can you run the tier selection procedure for that topic?"}],"resultDescription":"A structured response containing a prose decision procedure brief for agents, the recommended model tier, the topic, the service slug, an ISO-8601 timestamp, and the generation source (llm, reasoning, or deterministic). Guides the agent on how to select model tier, cap compute relative to expected value, enforce per-decision and per-hour budgets, detect tier creep, and run equivalence checks to justify downgrades.","failureModes":["Payment not included or insufficient (x402 payment required, returns 402)","Invalid query parameters (non-enum values for meta field)","Service unavailable or timeout from upstream provider","Empty or missing topic results in generic procedure rather than task-specific guidance","Handler failure indicated by ok: false in result object"],"whenToPreferThis":"Choose this endpoint when your agent needs a principled, cost-aware decision procedure for selecting LLM model tiers and enforcing inference budgets per decision — especially when tier creep, overspend, or uncontrolled frontier model usage is a concern. Prefer this over ad-hoc cost checks when you need a formalized gate that includes equivalence testing and value-to-compute ratio enforcement.","instructions":null,"reviewSummary":null,"reviewSummaryHighlights":null,"reviewSummaryConcerns":null,"reviewSummaryGeneratedAt":null,"activationCount":0,"lastUsedAt":null,"lastSuccessfullyRanAt":null,"lastHealthCheckAt":"2026-09-13T19:00:50.939Z","isFirstParty":false}