fix: update v2 session usage metrics (#35468)

packages/core/src/session/runner/llm.ts

@@ -17,6 +17,7 @@ import { Config } from "../../config"
 import { Database } from "../../database/database"
 import { EventV2 } from "../../event"
 import { Location } from "../../location"
+import { ModelV2 } from "../../model"
 import { PermissionV2 } from "../../permission"
 import { Instructions } from "../../instructions/index"
 import { InstructionBuiltIns } from "../../instructions/builtins"
@@ -50,6 +51,30 @@ import { StepFailedError, UserInterruptedError } from "../error"
 import { toSessionError } from "../to-session-error"
 import { SessionRunnerRetry } from "./retry"
 
+type StepTokens = {
+  readonly input: number
+  readonly output: number
+  readonly reasoning: number
+  readonly cache: { readonly read: number; readonly write: number }
+}
+
+// TODO(#35765): Use Copilot's reported billed amount once billing has a dedicated typed runtime contract.
+export function calculateCost(costs: ModelV2.Info["cost"], tokens: StepTokens) {
+  const context = tokens.input + tokens.cache.read + tokens.cache.write
+  const tier = costs
+    .filter((cost) => cost.tier?.type === "context" && context > cost.tier.size)
+    .toSorted((a, b) => (b.tier?.size ?? 0) - (a.tier?.size ?? 0))[0]
+  const cost = tier ?? costs.find((cost) => cost.tier === undefined)
+  if (!cost) return 0
+  return (
+    (tokens.input * cost.input +
+      (tokens.output + tokens.reasoning) * cost.output +
+      tokens.cache.read * cost.cache.read +
+      tokens.cache.write * cost.cache.write) /
+    1_000_000
+  )
+}
+
 /**
  * Runs one durable coding-agent Session until it settles.
  *
@@ -312,6 +337,11 @@ const layer = Layer.effect(
         Effect.ensuring(serialized(publisher.flush())),
       )
 
+      const stepUsage = (settlement: NonNullable<ReturnType<typeof publisher.stepSettlement>>) => ({
+        cost: calculateCost(resolved.cost, settlement.tokens),
+        tokens: settlement.tokens,
+      })
+
       // Captures the end snapshot, diffs it against the step's start, and durably ends the
       // assistant step.
       const publishStepEnd = (settlement: NonNullable<ReturnType<typeof publisher.stepSettlement>>) =>
@@ -328,8 +358,7 @@ const layer = Layer.effect(
               sessionID: session.id,
               assistantMessageID: yield* publisher.startAssistant(),
               finish: settlement.finish,
-              cost: 0,
-              tokens: settlement.tokens,
+              ...stepUsage(settlement),
               snapshot: endSnapshot,
               files,
             }),
@@ -452,7 +481,8 @@ const layer = Layer.effect(
           const stepEndedCleanly =
             !streamInterrupted && !toolsInterrupted && infraError === undefined && !providerFailed && !stepFailure
           if (stepSettlement && stepEndedCleanly) yield* publishStepEnd(stepSettlement)
-          if (stepFailure) yield* serialized(publisher.publishStepFailure())
+          if (stepFailure)
+            yield* serialized(publisher.publishStepFailure(stepSettlement ? stepUsage(stepSettlement) : undefined))
 
           if (stream._tag === "Failure") return yield* Effect.failCause(stream.cause)
           if (userDeclined) return yield* Effect.interrupt

packages/core/src/session/runner/model.ts

@@ -82,6 +82,8 @@ export interface Resolved {
   readonly model: Model
   /** Selected catalog identity. Durable records and displays must use this, never the API model id. */
   readonly ref: ModelV2.Ref
+  /** Catalog pricing in dollars per million tokens. */
+  readonly cost: ModelV2.Info["cost"]
 }
 
 export interface Interface {
@@ -94,13 +96,14 @@ export class Service extends Context.Service<Service, Interface>()("@opencode/v2
 export const layerWith = (resolve: Interface["resolve"]) => Layer.succeed(Service, Service.of({ resolve }))
 
 /** Builds a Resolved whose catalog identity mirrors the route model. Test or embedding seam. */
-export const resolved = (model: Model, variant?: ModelV2.VariantID): Resolved => ({
+export const resolved = (model: Model, variant?: ModelV2.VariantID, cost: ModelV2.Info["cost"] = []): Resolved => ({
   model,
   ref: ModelV2.Ref.make({
     id: ModelV2.ID.make(model.id),
     providerID: ProviderV2.ID.make(model.provider),
     ...(variant === undefined ? {} : { variant }),
   }),
+  cost,
 })
 
 const apiKey = (model: ModelV2.Info, credential?: Credential.Value) => {
@@ -341,6 +344,7 @@ const layer = Layer.effect(
             providerID: selected.providerID,
             ...(session.model?.variant === undefined ? {} : { variant: session.model.variant }),
           }),
+          cost: selected.cost,
         }
       }),
     })

packages/core/src/session/runner/publish-llm-event.ts

@@ -216,7 +216,10 @@ export const createLLMEventPublisher = (events: EventV2.Interface, input: Input)
     if (replace || stepFailure === undefined) stepFailure = error
   })
 
-  const publishStepFailure = Effect.fnUntraced(function* () {
+  const publishStepFailure = Effect.fnUntraced(function* (usage?: {
+    readonly cost: number
+    readonly tokens: ReturnType<typeof tokens>
+  }) {
     if (stepFailed || stepFailure === undefined) return
     const assistantMessageID = yield* startAssistant()
     stepFailed = true
@@ -224,6 +227,7 @@ export const createLLMEventPublisher = (events: EventV2.Interface, input: Input)
       sessionID: input.sessionID,
       assistantMessageID,
       error: stepFailure,
+      ...usage,
     })
   })
 
@@ -409,12 +413,12 @@ export const createLLMEventPublisher = (events: EventV2.Interface, input: Input)
       case "step-finish":
         yield* flush()
         if (stepSettlement) return yield* Effect.die(new Error("Duplicate step finish"))
+        stepSettlement = { finish: event.reason, tokens: tokens(event.usage) }
         if (event.reason === "content-filter") {
           providerFailed = true
           yield* failAssistant({ type: "provider.content-filter", message: "Provider blocked the response" }, true)
           return
         }
-        stepSettlement = { finish: event.reason, tokens: tokens(event.usage) }
         return
       case "finish":
         return