@oh-my-pi/pi-catalog 18.3.5 → 18.4.0

package/CHANGELOG.md

@@ -2,6 +2,12 @@
 
 ## [Unreleased]
 
+## [18.4.0] - 2026-09-28
+
+### Fixed
+
+- Corrected Cerebras model pricing so cached input tokens are charged at the model’s input rate, consistent with Cerebras billing.
+
 ## [18.3.5] - 2026-09-27
 
 ### Added

package/dist/types/build.d.ts

@@ -9,8 +9,8 @@
 import { type Api, type Model, type ModelSpec } from "./types.js";
 /**
  * Applies reviewed catalog-data value corrections (`cost-patch`,
- * `limits-patch`, `long-context-cost`, `context-window-floor`,
- * `input-modalities`) onto an upstream-sourced spec. Applied by
+ * `cache-read-at-input-rate`, `limits-patch`, `long-context-cost`,
+ * `context-window-floor`, `input-modalities`) onto an upstream-sourced spec. Applied by
  * `buildModel` to every upstream-sourced spec; user-authored overrides are
  * recomposed after building by the override applicators, so explicit user
  * limits and pricing still win.

package/dist/types/discovery/cursor-proto.d.ts

@@ -5070,6 +5070,11 @@ export interface TruncatedToolCallSuccess extends ProtoMessage {
 export declare const TruncatedToolCallSuccessSchema: MessageCodec<TruncatedToolCallSuccess>;
 /** Cursor agent message agent.v1.TurnEndedUpdate. */
 export interface TurnEndedUpdate extends ProtoMessage {
+    inputTokens?: bigint;
+    outputTokens?: bigint;
+    cacheReadTokens?: bigint;
+    cacheWriteTokens?: bigint;
+    reasoningTokens?: bigint;
 }
 export declare const TurnEndedUpdateSchema: MessageCodec<TurnEndedUpdate>;
 /** Cursor agent message agent.v1.TypeAction. */

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@oh-my-pi/pi-catalog",
-	"version": "18.3.5",
+	"version": "18.4.0",
 	"description": "Model catalog for omp: bundled model database, provider discovery descriptors, model identity, classification, and equivalence",
 	"keywords": [
 		"ai",
@@ -93,12 +93,12 @@
 		"gen:proto": "bun scripts/generate-protocols.ts"
 	},
 	"dependencies": {
-		"@oh-my-pi/omptype": "18.3.5",
-		"@oh-my-pi/pi-utils": "18.3.5"
+		"@oh-my-pi/omptype": "18.4.0",
+		"@oh-my-pi/pi-utils": "18.4.0"
 	},
 	"devDependencies": {
 		"@bgotink/kdl": "0.4.0",
-		"@oh-my-pi/pi-ai": "18.3.5",
+		"@oh-my-pi/pi-ai": "18.4.0",
 		"@types/bun": "^1.3.14"
 	},
 	"engines": {

package/src/build.ts

@@ -140,8 +140,8 @@ function applyCatalogAssignments<TApi extends Api>(model: Model<TApi>, catalog:
 
 /**
  * Applies reviewed catalog-data value corrections (`cost-patch`,
- * `limits-patch`, `long-context-cost`, `context-window-floor`,
- * `input-modalities`) onto an upstream-sourced spec. Applied by
+ * `cache-read-at-input-rate`, `limits-patch`, `long-context-cost`,
+ * `context-window-floor`, `input-modalities`) onto an upstream-sourced spec. Applied by
  * `buildModel` to every upstream-sourced spec; user-authored overrides are
  * recomposed after building by the override applicators, so explicit user
  * limits and pricing still win.
@@ -220,6 +220,14 @@ export function applyCatalogCorrections(
 			applyEffectiveFallbackRates(model.cost, Reflect.get(fallback, "effectiveRates"));
 		}
 	}
+	if (catalog.cacheReadAtInputRate === true) {
+		const { longContext } = model.cost;
+		model.cost = {
+			...model.cost,
+			cacheRead: model.cost.input,
+			...(longContext && { longContext: { ...longContext, cacheRead: longContext.input } }),
+		};
+	}
 	if (catalog.timeBased !== undefined) {
 		model.cost = { ...model.cost, timeBased: materializeTimeBasedCost(catalog.timeBased) };
 	}

package/src/compat/axes.ts

@@ -321,6 +321,16 @@ export const AXES: Readonly<Record<string, AxisDef>> = {
 	"context-window-floor": { key: "contextWindowFloor", set: "catalog", shape: "scalar" },
 	"cost-patch": { key: "costPatch", set: "catalog", shape: "object" },
 	"cost-fallback": { key: "costFallback", set: "catalog", shape: "object" },
+	/**
+	 * The host bills cache-hit input tokens at the full input rate (no cache
+	 * discount), so the built row's `cacheRead` tracks its live `input` price.
+	 */
+	"cache-read-at-input-rate": {
+		key: "cacheReadAtInputRate",
+		set: "catalog",
+		shape: "scalar",
+		values: [true, false],
+	},
 	"delegation-bias": { key: "delegationBias", set: "catalog", shape: "scalar", values: DELEGATION_BIASES },
 	"discovery-api": { key: "discoveryApi", set: "catalog", shape: "scalar" },
 	"edit-prompt-variant": { key: "editPromptVariant", set: "catalog", shape: "scalar", values: ["full", "compact"] },

package/src/compat/rules/providers/cerebras.kdl

@@ -10,6 +10,9 @@ provider "cerebras" {
 	// Replaces the Cerebras provider branches in the completions baseline.
 	supports-usage-in-streaming #false
 	tool-strict-mode "all_strict"
+	// Prompt caching carries no discount: cached input tokens bill at the
+	// standard input rate (https://inference-docs.cerebras.ai/capabilities/prompt-caching).
+	cache-read-at-input-rate #true
 	// Cerebras serves Gemma 4 with vision enabled; its /v1/models rows carry
 	// no modality metadata.
 	models "gemma-4-31b" {