@oh-my-pi/pi-catalog 16.3.14 → 16.4.0

package/CHANGELOG.md

@@ -2,6 +2,43 @@
 
 ## [Unreleased]
 
+## [16.4.0] - 2026-07-10
+
+### Breaking Changes
+
+- Redesigned reasoning effort ladders to be wire-exact, removing the shifted five-tier effort mapping. Models now expose exactly the effort tiers their upstream APIs accept, mapped 1:1. Removed SHIFTED_FIVE_TIER_EFFORT_MAP, ANTHROPIC_ADAPTIVE_EFFORT_MAP_4_TIER, and per-host xhigh-to-max alias maps. Selecting an unsupported tier now automatically clamps down via clampThinkingLevelForModel. Devin effort routing is now mapped 1:1 onto per-tier siblings.
+
+### Added
+
+- Added support for new models: Grok 4.5 family, Dolphin Mistral 24b Venice Edition, GLM5.2-Fast, and Zenmux variants for GPT-5.6 (Luna, Sol, and Terra).
+- Added Novita as a model provider, including public catalog discovery, pricing, limits, modality, reasoning, and tool metadata.
+- Added useResponsesLite to Model and ModelSpec to support the Responses Lite transport, enabled by default for the GPT-5.6 family.
+- Added Effort.Max ("max") as a first-class user-facing thinking level above xhigh.
+
+### Changed
+
+- Enabled reasoning effort controls for Grok 4.5 and updated support flags for additional Grok variants
+- Standardized reasoning effort levels to use a wire-exact max tier across all model providers, including Devin routing and Ollama configurations.
+- Updated costs and context windows for various models in the catalog.
+
+## [16.3.15] - 2026-07-09
+
+### Added
+
+- Added support for Grok 4.5 model
+- Added `gpt-5.6` base models and `gpt-5.6-{luna,sol,terra}-pro` variants
+- Added `meta/muse-spark-1.1` model support
+- Added support for thinking modes on `poolside/laguna` models
+- Added generated GPT-5.6 Pro aliases (`gpt-5.6-{luna,sol,terra}-pro`) on the `openai` and `openai-codex` providers: each alias sends the base model id on the wire (`requestModelId`) with the new `reasoningMode: "pro"` marker, and re-derives from the current base rows on every catalog regeneration.
+
+### Changed
+
+- Updated cache read costs for Grok models
+- Reduced max token limit for Grok 4.3 model
+- Enabled prompt cache affinity for Grok models via the x-grok-conv-id header in OpenAI compatible endpoints
+- Enabled prompt cache affinity for Grok models via the x-grok-conv-id header
+- Marked direct xAI Grok Chat Completions models for `x-grok-conv-id` prompt-cache affinity.
+
 ## [16.3.14] - 2026-07-09
 
 ### Added

package/dist/types/effort.d.ts

@@ -4,6 +4,7 @@ export declare const enum Effort {
     Low = "low",
     Medium = "medium",
     High = "high",
-    XHigh = "xhigh"
+    XHigh = "xhigh",
+    Max = "max"
 }
 export declare const THINKING_EFFORTS: readonly Effort[];

package/dist/types/model-thinking.d.ts

@@ -13,21 +13,6 @@ import type { Api, CompatOf, Model, ModelSpec, ThinkingConfig } from "./types.js
  * specs and built models.
  */
 type ApiModel<TApi extends Api = Api> = ModelSpec<TApi> | Model<TApi>;
-/**
- * Effort → wire-value map for a shifted five-tier scale (`low..max`):
- * user-facing efforts shift up one notch so the top tier reaches the genuine
- * "max" and "high" lands on the recommended "xhigh" coding/agentic default.
- * Used by Anthropic adaptive models with a real xhigh tier (Opus 4.7+ and
- * Fable/Mythos 5 on the Messages API) and by GPT-5.6+ wire-effort models,
- * which expose the same genuine `max` tier above `xhigh`.
- */
-export declare const SHIFTED_FIVE_TIER_EFFORT_MAP: Readonly<Partial<Record<Effort, string>>>;
-/**
- * Effort → wire-value map for the legacy 4-tier adaptive scale (Opus 4.6,
- * Sonnet 4.6+, and every adaptive model on Bedrock Converse). `low..high` pass
- * through verbatim; there is no real "xhigh", so it aliases the top "max" tier.
- */
-export declare const ANTHROPIC_ADAPTIVE_EFFORT_MAP_4_TIER: Readonly<Partial<Record<Effort, string>>>;
 /**
  * Resolve the canonical thinking metadata for a spec. Called exactly once per
  * model by `buildModel`, after compat resolution.

package/dist/types/provider-models/descriptors.d.ts

@@ -224,6 +224,16 @@ export declare const CATALOG_PROVIDERS: readonly [{
     readonly catalogDiscovery: {
         readonly label: "NVIDIA";
     };
+}, {
+    readonly id: "novita";
+    readonly defaultModel: "moonshotai/kimi-k2.7-code";
+    readonly envVars: readonly ["NOVITA_API_KEY"];
+    readonly createModelManagerOptions: (config: ModelManagerConfig) => import("../index.js").ModelManagerOptions<"openai-completions", unknown>;
+    readonly dynamicModelsAuthoritative: true;
+    readonly catalogDiscovery: {
+        readonly label: "Novita";
+        readonly allowUnauthenticated: true;
+    };
 }, {
     readonly id: "ollama";
     readonly defaultModel: "gpt-oss:20b";

package/dist/types/provider-models/openai-compat.d.ts

@@ -54,6 +54,18 @@ export interface OpenAIModelManagerConfig {
     fetch?: FetchImpl;
 }
 export declare function openaiModelManagerOptions(config?: OpenAIModelManagerConfig): ModelManagerOptions<"openai-responses">;
+/**
+ * Re-derive the generated pro-reasoning aliases (`gpt-5.6-*-pro`) for the
+ * first-party `openai`/`openai-codex` gpt-5.6 rows. Each alias inherits the
+ * base row's metadata, requests the base wire id via `requestModelId`, and
+ * sets `reasoningMode: "pro"` so Responses-family request builders emit
+ * `reasoning: { mode: "pro" }`. Called by the models.json generator after all
+ * sources merge: stale copies of the owned aliases (previous snapshot) are
+ * dropped and re-projected from the current base rows so alias metadata always
+ * tracks the base, while a real upstream model that occupies an alias id wins
+ * and suppresses the projection.
+ */
+export declare function projectOpenAIProReasoningAliases(models: readonly ModelSpec<Api>[]): ModelSpec<Api>[];
 export interface GroqModelManagerConfig {
     apiKey?: string;
     baseUrl?: string;
@@ -78,6 +90,14 @@ export interface NvidiaModelManagerConfig {
     fetch?: FetchImpl;
 }
 export declare function nvidiaModelManagerOptions(config?: NvidiaModelManagerConfig): ModelManagerOptions<"openai-completions">;
+/** Novita OpenAI-compatible discovery configuration. */
+export interface NovitaModelManagerConfig {
+    apiKey?: string;
+    baseUrl?: string;
+    fetch?: FetchImpl;
+}
+/** Builds Novita's public model-discovery manager. */
+export declare function novitaModelManagerOptions(config?: NovitaModelManagerConfig): ModelManagerOptions<"openai-completions">;
 export interface XaiModelManagerConfig {
     apiKey?: string;
     baseUrl?: string;

package/dist/types/types.d.ts

@@ -563,6 +563,13 @@ export interface Model<TApi extends Api = Api> {
      * everything local (selection, caching, usage attribution) keys on `id`.
      */
     requestModelId?: string;
+    /**
+     * `reasoning.mode` to send on OpenAI Responses-family requests. Set on
+     * generated pro aliases (`gpt-5.6-*-pro` on `openai`/`openai-codex`) that
+     * pair a base wire id (`requestModelId`) with OpenAI's pro reasoning
+     * serving path. Absent everywhere else; providers omit the wire field.
+     */
+    reasoningMode?: "pro";
     name: string;
     api: TApi;
     provider: Provider;
@@ -623,6 +630,8 @@ export interface Model<TApi extends Api = Api> {
     transport?: "pi-native";
     /** Hint that websocket transport should be preferred when supported by the provider implementation. */
     preferWebsockets?: boolean;
+    /** Codex Responses Lite transport: send the lite marker and carry instructions/tools as input items (mirrors codex-rs `use_responses_lite`). */
+    useResponsesLite?: boolean;
     /** Preferred model to switch to when context promotion is triggered (model id or provider/id). */
     contextPromotionTarget?: string;
     /** Preferred model to use only for compaction (model id or provider/id); the active session model is unchanged. */

package/dist/types/wire/codex.d.ts

@@ -8,6 +8,15 @@ export declare const OPENAI_HEADERS: {
     readonly ORIGINATOR: "originator";
     readonly SESSION_ID: "session_id";
     readonly CONVERSATION_ID: "conversation_id";
+    readonly SCOPED_SESSION_ID: "session-id";
+    readonly THREAD_ID: "thread-id";
+    readonly INSTALLATION_ID: "x-codex-installation-id";
+    readonly WINDOW_ID: "x-codex-window-id";
+    readonly TURN_METADATA: "x-codex-turn-metadata";
+    readonly PARENT_THREAD_ID: "x-codex-parent-thread-id";
+    readonly SUBAGENT: "x-openai-subagent";
+    /** Responses Lite transport marker (codex-rs `add_responses_lite_header`); value is always `"true"`. */
+    readonly RESPONSES_LITE: "x-openai-internal-codex-responses-lite";
 };
 export declare const OPENAI_HEADER_VALUES: {
     readonly BETA_RESPONSES: "responses=experimental";

package/package.json

@@ -1,7 +1,7 @@
 {
 	"type": "module",
 	"name": "@oh-my-pi/pi-catalog",
-	"version": "16.3.14",
+	"version": "16.4.0",
 	"description": "Model catalog for omp: bundled model database, provider discovery descriptors, model identity, classification, and equivalence",
 	"homepage": "https://omp.sh",
 	"author": "Can Boluk",
@@ -34,12 +34,12 @@
 	},
 	"dependencies": {
 		"@bufbuild/protobuf": "^2.12.0",
-		"@oh-my-pi/pi-utils": "16.3.14",
+		"@oh-my-pi/pi-utils": "16.4.0",
 		"arktype": "^2.2.0",
 		"zod": "^4"
 	},
 	"devDependencies": {
-		"@oh-my-pi/pi-ai": "16.3.14",
+		"@oh-my-pi/pi-ai": "16.4.0",
 		"@types/bun": "^1.3.14"
 	},
 	"engines": {

package/src/compat/openai.ts

@@ -71,31 +71,11 @@ const DSML_HEALING_PROVIDERS = new Set([
 	"openrouter",
 ]);
 
-/**
- * Ollama's OpenAI-compatible `reasoning.effort` only accepts
- * `high|medium|low|max|none`; OMP's `minimal`/`xhigh` levels make the server
- * reject the turn with HTTP 400 `invalid reasoning value`. Map the two
- * unsupported levels onto the closest accepted ones. Stamped in the compat
- * builder (not only at discovery) so stale-cached and custom `ollama`-provider
- * specs are backfilled on every `buildModel`, not just on a fresh
- * `omp models refresh`. Custom OpenAI-compatible providers pointed at a local
- * Ollama port under a different provider id are not covered — they must set
- * `compat.reasoningEffortMap` themselves.
- */
-const OLLAMA_REASONING_EFFORT_MAP: ResolvedOpenAISharedCompat["reasoningEffortMap"] = { minimal: "low", xhigh: "max" };
-
-/**
- * Merge the Ollama default effort map under any explicit overrides (overrides
- * win). No-op off the local `ollama` provider or for non-reasoning models.
- */
-function mergeOllamaReasoningEffortMap(
-	compat: ResolvedOpenAISharedCompat,
-	provider: string,
-	reasoning: boolean | undefined,
-): void {
-	if (provider !== "ollama" || !reasoning) return;
-	compat.reasoningEffortMap = { ...OLLAMA_REASONING_EFFORT_MAP, ...compat.reasoningEffortMap };
-}
+// Ollama's OpenAI-compatible `reasoning.effort` accepts `high|medium|low|max|none`;
+// `ollama`-provider reasoning models carry the wire-exact `low..max` effort
+// ladder (see getModelDefinedEfforts), so no compat-level remapping is needed.
+// Custom OpenAI-compatible providers pointed at a local Ollama port under a
+// different provider id must set `compat.reasoningEffortMap` themselves.
 
 function resolveReasoningDisableMode(
 	thinkingFormat: ResolvedOpenAISharedCompat["thinkingFormat"],
@@ -541,7 +521,7 @@ export function buildOpenAICompat(spec: ModelSpec<"openai-completions">): Resolv
 			MINIMAX_PROVIDER_OR_ID_PATTERN.test(provider) || MINIMAX_PROVIDER_OR_ID_PATTERN.test(spec.id),
 		emptyLengthFinishIsContextError: provider === "ollama",
 		usesOpenAIToolCallIdLimit: provider === "openai",
-		promptCacheSessionHeader: undefined,
+		promptCacheSessionHeader: isGrok ? "x-grok-conv-id" : undefined,
 		dropThinkingWhenReasoningEffort: provider === "fireworks",
 	};
 
@@ -554,7 +534,6 @@ export function buildOpenAICompat(spec: ModelSpec<"openai-completions">): Resolv
 	if (spec.compat?.omitReasoningEffort === undefined && !compat.supportsReasoningEffort) {
 		compat.omitReasoningEffort = true;
 	}
-	mergeOllamaReasoningEffortMap(compat, provider, spec.reasoning);
 	mergeMimoReasoningEffortMap(compat, isMimoReasoningEffortModel);
 
 	const whenThinkingPolicy =
@@ -568,7 +547,6 @@ export function buildOpenAICompat(spec: ModelSpec<"openai-completions">): Resolv
 		if (whenThinkingPolicy.omitReasoningEffort === undefined && !variant.supportsReasoningEffort) {
 			variant.omitReasoningEffort = true;
 		}
-		mergeOllamaReasoningEffortMap(variant, provider, spec.reasoning);
 		mergeMimoReasoningEffortMap(variant, isMimoReasoningEffortModel);
 		compat.whenThinking = variant;
 	}
@@ -678,7 +656,6 @@ export function buildOpenAIResponsesCompat(spec: OpenAIResponsesSpecLike): Resol
 	if (spec.compat?.omitReasoningEffort === undefined && !compat.supportsReasoningEffort) {
 		compat.omitReasoningEffort = true;
 	}
-	mergeOllamaReasoningEffortMap(compat, spec.provider, spec.reasoning);
 	return compat;
 }
 

package/src/discovery/codex.ts

@@ -30,6 +30,7 @@ const codexModelEntrySchema = type({
 	"supported_in_api?": "unknown",
 	"priority?": "unknown",
 	"prefer_websockets?": "unknown",
+	"use_responses_lite?": "unknown",
 });
 
 const codexModelsResponseSchema = type({
@@ -262,6 +263,7 @@ function normalizeCodexModelEntry(entry: unknown, baseUrl: string): NormalizedCo
 	const reasoning = supportsReasoning(payload.default_reasoning_level, payload.supported_reasoning_levels);
 	const input = normalizeInputModalities(payload.input_modalities);
 	const preferWebsockets = toBoolean(payload.prefer_websockets) === true;
+	const useResponsesLite = toBoolean(payload.use_responses_lite) === true;
 	const priority = toFiniteNumber(payload.priority) ?? Number.MAX_SAFE_INTEGER;
 
 	return {
@@ -279,6 +281,7 @@ function normalizeCodexModelEntry(entry: unknown, baseUrl: string): NormalizedCo
 			contextWindow,
 			maxTokens,
 			...(preferWebsockets ? { preferWebsockets: true } : {}),
+			...(useResponsesLite ? { useResponsesLite: true } : {}),
 			...(priority !== Number.MAX_SAFE_INTEGER ? { priority } : {}),
 		},
 	};

package/src/effort.ts

@@ -5,6 +5,7 @@ export const enum Effort {
 	Medium = "medium",
 	High = "high",
 	XHigh = "xhigh",
+	Max = "max",
 }
 
 export const THINKING_EFFORTS: readonly Effort[] = [
@@ -13,4 +14,5 @@ export const THINKING_EFFORTS: readonly Effort[] = [
 	Effort.Medium,
 	Effort.High,
 	Effort.XHigh,
+	Effort.Max,
 ];

package/src/identity/family.ts

@@ -78,7 +78,7 @@ export const isMimoModelIdOrName = memo((value: string): boolean => {
 	return value.toLowerCase().includes("mimo");
 });
 
-const GROK_EFFORT_CAPABLE_PREFIXES = ["grok-3-mini", "grok-4.20-multi-agent", "grok-4.3"] as const;
+const GROK_EFFORT_CAPABLE_PREFIXES = ["grok-3-mini", "grok-4.20-multi-agent", "grok-4.3", "grok-4.5"] as const;
 
 /**
  * Grok SKUs that expose the wire `reasoning.effort` dial. Other Grok reasoners