> ## Documentation Index > Fetch the complete documentation index at: https://docs.sqwish.ai/llms.txt > Use this file to discover all available pages before exploring further. # Estimate the fee of starting a family > The exact fee `POST /v1/fine-tuning/jobs` would hold for the same body now, without max_price_usd, and the training tokens and steps it is for. It reads and counts the dataset as the start does, and refuses what the start would about the dataset, the family name, the base, the rows and the steps. Nothing is queued or held, and credit, storage and queue room are checked when the job starts. The estimate is exact if nothing changes before you start; pass it as max_price_usd to cap the price. ## OpenAPI ````yaml /openapi.json post /v1/fine-tuning/jobs/estimate openapi: 3.1.0 info: title: DecisionOne workbench description: >- Native decisions, versioned datasets, adaptation and shared-base deployment. Authenticate with `Authorization: Bearer d1_sk_...` when accounts are on. A short guide for coding agents is at /llms.txt. Hosted account APIs accept a Bearer API key or a browser session. An API key reads its account but can't change it: keys, billing and settings change in the console. Browser mutations require the configured Origin and X-CSRF-Token from /auth/session. Account recovery and history remain available to blocked accounts. Waitlisted accounts can read their account summary and model catalogue and request early access; private member APIs require admission. Money fields ending in _usd are exact decimal strings; never interpret them as integer cents. Authenticated customer reads and existing-operation replays share read_rpm, and other customer writes except decisions and their estimates share write_rpm. Exhausted budgets return 429 with Retry-After; unavailable admission storage returns 503. version: 0.1.0 servers: - url: https://console.sqwish.ai security: [] tags: - name: Decisions description: Probabilities for questions with fixed outcomes. - name: Models description: Base models, and fine-tuned models by their numbers. - name: Families description: >- Lines of models trained on one dataset, and the switches that let them answer calls. - name: Datasets description: >- Validated training examples. A dataset grows in versions, and a version never changes. - name: Evaluations description: Reusable model results on a dataset's fixed evaluation rows. - name: Fine-tuning description: Training jobs, their events and held-out results. - name: Prompt tuning description: Better wording for a decision, tested on your rows. Words, not weights. - name: Labeling description: Teacher labels for real queries, reviewed by you. - name: Accounts description: Email OTP sign-in, revocable sessions, security and API keys. - name: Billing description: Exact USD balances, purchases, automatic top-ups and referrals. - name: Integrations description: Endpoints other tools call, such as Claude Code. - name: System description: Health and discovery. paths: /v1/fine-tuning/jobs/estimate: post: tags: - Fine-tuning summary: Estimate the fee of starting a family description: >- The exact fee `POST /v1/fine-tuning/jobs` would hold for the same body now, without max_price_usd, and the training tokens and steps it is for. It reads and counts the dataset as the start does, and refuses what the start would about the dataset, the family name, the base, the rows and the steps. Nothing is queued or held, and credit, storage and queue room are checked when the job starts. The estimate is exact if nothing changes before you start; pass it as max_price_usd to cap the price. operationId: estimateFineTuningJob requestBody: content: application/json: schema: $ref: '#/components/schemas/JobEstimate' required: true responses: '200': description: Successful Response content: application/json: schema: properties: currency: type: string const: USD title: Currency price_usd: description: >- The fee a fine-tune started now with the same body would hold, as exact USD: its base's fine_tune_million_tokens_usd for trained_tokens, and at least fine_tune_minimum_usd. 0 when it would be free. The job's price_usd is this sum while the prices, the account, the dataset and the family stay as they are; send it as max_price_usd, and a job that would cost more is refused with price_changed. title: Price Usd type: string minimum_applied: description: >- True when the fee is fine_tune_minimum_usd, which is more than its training tokens cost. title: Minimum Applied type: boolean trained_tokens: anyOf: - type: integer - type: 'null' description: >- The training tokens it would train, as the job's split.trained_tokens counts them. Null only for a free fine-tune whose rows can't be counted. title: Trained Tokens steps: description: 'Its training steps: the ones sent, or the default passes.' title: Steps type: integer seed: description: Its seed, which orders the rows its steps take. title: Seed type: integer passes: description: >- How many times its steps go over its training rows, to two decimals. title: Passes type: number train_rows: description: >- The training rows it trains on: a continuation's new rows and the earlier rows sampled with them. title: Train Rows type: integer dataset_version: description: 'The dataset version it would train on: its newest.' title: Dataset Version type: integer required: - currency - price_usd - minimum_applied - trained_tokens - steps - seed - passes - train_rows - dataset_version title: FineTuneEstimate type: object 4XX: description: >- Client error. On a 422, `error.details` lists each field that failed. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' 5XX: description: Server error. A 503 names what is not running on this server. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' security: - BearerAuth: [] components: schemas: JobEstimate: properties: dataset_id: type: string pattern: ^ds_[a-f0-9]{24}$ title: Dataset Id from: type: string maxLength: 140 minLength: 1 title: From description: >- Where the family starts: a base model that can be fine-tuned, or one of your models as family@N, which it trains from on that model's base. Left out, Core. default: sqwish-decision-core family: type: string maxLength: 63 pattern: ^[a-z](?:-?[a-z0-9]){1,62}$ title: Family description: >- The new family's name: lowercase letters and digits, with single hyphens between them, 2 to 63 characters. A name already taken is refused with family_exists. method: type: string enum: - sft - reward title: Method default: sft hyperparameters: $ref: '#/components/schemas/Hyperparameters' additionalProperties: false type: object required: - dataset_id - family title: JobEstimate description: >- A new family's first fine-tune, from a base or one of your models, as its estimate takes it: the body that starts it, without max_price_usd. ErrorResponse: description: >- The body of every error. app.failure() builds it; this model documents it. properties: error: $ref: '#/components/schemas/ErrorBody' required: - error title: ErrorResponse type: object Hyperparameters: properties: steps: anyOf: - type: integer maximum: 2000 minimum: 1 - type: 'null' title: Steps description: >- Training steps. Left out, two passes over the training rows, from 20 to 2,000 steps, so a very large dataset gets fewer than two passes. At most the policy's limits.fine_tune_passes passes (2 by default), or 20 steps. The fee is per training token trained, so more steps train more tokens and cost more. learning_rate: type: number maximum: 0.01 exclusiveMinimum: 0 title: Learning Rate default: 0.00005 lora_rank: type: integer minimum: 1 title: Lora Rank description: >- The LoRA rank of the adapter, from 1 to 16, the highest at launch. Left out, 8. A rank above 16 is refused with `rank_unsupported` when the job is accepted. A continuation keeps its parent's rank. default: 8 seed: type: integer maximum: 2147483647 minimum: 0 title: Seed default: 42 checkpoint_every: type: integer maximum: 1000 minimum: 1 title: Checkpoint Every default: 10 kl_coefficient: type: number maximum: 10 minimum: 0 title: Kl Coefficient default: 0.05 additionalProperties: false type: object title: Hyperparameters ErrorBody: properties: code: description: A stable code to branch on, such as invalid_request. title: Code type: string message: description: What went wrong and what to change, in plain words. title: Message type: string retryable: description: >- True when another attempt may succeed; it does not prove that no work was done. Respect Retry-After when present. Only operations that support Idempotency-Key replay a prior result, including /v1/decide for 24 hours. title: Retryable type: boolean request_id: description: Also sent as X-Request-ID. Quote it in bug reports. title: Request Id type: string details: anyOf: - items: $ref: '#/components/schemas/ErrorDetail' type: array - type: 'null' default: null description: On validation errors, one entry for each field that failed. title: Details required: - code - message - retryable - request_id title: ErrorBody type: object ErrorDetail: properties: path: description: Where the problem is, such as ["body", "decisions", 0, "outcomes"]. items: anyOf: - type: string - type: integer title: Path type: array message: title: Message type: string type: anyOf: - type: string - type: 'null' default: null title: Type required: - path - message title: ErrorDetail type: object securitySchemes: BearerAuth: type: http scheme: bearer bearerFormat: d1_sk_... description: Create an API key in the console. Required when accounts are enabled. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.