Sample data. These figures are not a customer bill.Learn more
parezenparezen
Northwind AI
Scale · Sample · cloud
Home
  • Get startedSoon
  • Audits
  • Overview
Analyse
  • Spend
  • Cost per task
  • Agents
  • Model comparisonSoon
  • Budgets
  • AlertsSoon
  • Reports
  • PR cost checksSoon
  • AdvisorsSoon
  • Spend analysis
Optimise
  • RecommendationsSoon
  • Routing policiesSoon
  • Shadow testsSoon
  • CachingSoon
  • TrimmingSoon
  • BatchSoon
  • Waste detectorSoon
Gateway
  • SetupSoon
  • ProvidersSoon
  • ModelsSoon
  • API keysSoon
  • DeploymentsSoon
  • PlaygroundSoon
  • LogsSoon
Savings & billing
  • Savings history
  • Savings ledger
  • Bill match
  • Billing
  • Invoices
Settings
  • Data & privacySoon
  • OrganisationSoon
  • ProjectsSoon
  • TeamSoon
  • RolesSoon
  • SSOSoon
  • Audit logSoon
Plan: Scale · Sample ↗
DocsHelp

Sample data. Not a customer bill.

Northwind AI/Advisors/Fine-tune advisor

Fine-tune advisor

Planned

The one high-volume task where a tuned small model might pay off.

Figures follow the shell: Last 30 days · All projects · All environments.

Shadow test required
Lever 11 cannot ship until a shadow test of the candidate model passes. Right-size tests on the current route do not count.
Task
ticket resolved
13,209 requests
Spend on this task
$20.4K$20.4K
Estimated opportunity
$19.3Kest.
Catalogue price ratio. Not verified.
Shadow test
Not passed

How the estimate is sized

It applies the catalogue output-price ratio to this task's spend. It is not a forecast and it is not billed.

Served now
Claude Haiku 4.5 · $5.00 / MTok output
Cheapest smaller catalogue model
Llama 3.3 70B · $0.28 / MTok output
Start a shadow testExisting shadow tests

Related shadow tests

Shown so a passed right-size test is not mistaken for a fine-tune test.

  • Haiku for support replyPassed · anthropic/claude-haiku-4.5
  • Flash for repeated support macrosNot passed · google/gemini-2.5-flash