@yansigit/opencodex 2.33.0 → 2.35.0

package/src/usage/expected-prices.ts

@@ -39,7 +39,7 @@ export interface ExpectedPriceOverlay {
 }
 
 const GEMINI_31_PRO: Cost4 = { input: 2, output: 12, cacheRead: 0.2, cacheWrite: 0 };
-const GPT56_SOL: Cost4 = { input: 5, output: 30, cacheRead: 0.5, cacheWrite: 6.25 };
+const GPT56_SOL: Cost4 = { input: 4, output: 20, cacheRead: 0.4, cacheWrite: 5 };
 /**
  * Daybreak aliases. `daybreak-*-latest` never appears in the pricing table itself — only its
  * current snapshot does — so these tuples are the snapshot's published rates and carry
@@ -89,6 +89,17 @@ const KIMI_PRICING = "https://platform.kimi.ai/docs/pricing (official table; cac
 // anywhere. Cache stays 0 rather than inheriting the reseller's 0.15 — a reseller number
 // under a vendor-price label would be a wrong value wearing a verified badge.
 const QWEN38_MAX_PRICING = "https://qwen.ai/blog?id=qwen3.8 (Qwen release announcement; no Model Studio billing row yet; cache rates unpublished -> 0)";
+const COMMAND_CODE_PRICING = "https://commandcode.ai/models + https://commandcode.ai/models.data (official model manifest)";
+const COMMAND_CODE_VERIFIED_AT = "2026-08-26";
+const CURSOR_DOCS_PRICING = "https://cursor.com/docs/models-and-pricing";
+const CURSOR_GROK_46: Cost4 = { input: 2, output: 6, cacheRead: 0.5, cacheWrite: 0 };
+const CURSOR_GROK_45: Cost4 = { input: 2, output: 6, cacheRead: 0.5, cacheWrite: 0 };
+const CURSOR_COMPOSER_1: Cost4 = { input: 1.25, output: 10, cacheRead: 0.125, cacheWrite: 0 };
+const CURSOR_COMPOSER_25: Cost4 = { input: 0.5, output: 2.5, cacheRead: 0.2, cacheWrite: 0 };
+const CURSOR_COMPOSER_25_FAST: Cost4 = { input: 3, output: 15, cacheRead: 0.5, cacheWrite: 0 };
+const CURSOR_GROK_46_FAST: Cost4 = { input: 4, output: 12, cacheRead: 1, cacheWrite: 0 };
+const CURSOR_GROK_45_FAST: Cost4 = { input: 4, output: 18, cacheRead: 1, cacheWrite: 0 };
+const CLAUDE_OPUS_FAST: Cost4 = { input: 10, output: 50, cacheRead: 1, cacheWrite: 12.5 };
 
 export const EXPECTED_PRICE_OVERLAYS: readonly ExpectedPriceOverlay[] = [
   // claude-opus-5 is exposed by three providers but absent from the jawcode bundle, so
@@ -126,6 +137,10 @@ export const EXPECTED_PRICE_OVERLAYS: readonly ExpectedPriceOverlay[] = [
   // Daybreak aliases: priced as their current snapshots (red -> gpt-5.6-cyber,
   // blue -> gpt-5.6-sol). The alias ids carry no rows of their own upstream, hence
   // verified-derived. Blue deliberately reuses GPT56_SOL rather than duplicating the tuple.
+  // The `gpt-` selector is native to ChatGPT/Codex accounts. The bare aliases below belong to
+  // the separately billed API-key catalog; keep those namespaces disjoint so pricing metadata
+  // cannot make an unroutable provider/model pair look supported.
+  { provider: "openai", modelId: "gpt-daybreak-blue-latest", cost4: GPT56_SOL, source: `alias of gpt-5.6-sol ${OPENAI_GPT56_PRICING}`, verifiedAt: "2026-08-11", status: "verified-derived" },
   { provider: "openai-apikey", modelId: "daybreak-red-latest", cost4: DAYBREAK_RED, source: `alias of gpt-5.6-cyber ${OPENAI_GPT56_PRICING}`, verifiedAt: "2026-08-11", status: "verified-derived" },
   { provider: "openai-apikey", modelId: "daybreak-blue-latest", cost4: GPT56_SOL, source: `alias of gpt-5.6-sol ${OPENAI_GPT56_PRICING}`, verifiedAt: "2026-08-11", status: "verified-derived" },
   { provider: "google-antigravity", modelId: "gemini-3.1-pro-low", cost4: GEMINI_31_PRO, source: `derived: gemini-3.1-pro (<=200k tier) ${GEMINI_PRICING}`, verifiedAt: "2026-07-20", status: "verified-derived" },
@@ -179,6 +194,16 @@ export const EXPECTED_PRICE_OVERLAYS: readonly ExpectedPriceOverlay[] = [
   { provider: "alibaba-token-plan-intl", modelId: "qwen3.8-max", cost4: QWEN38_MAX, source: QWEN38_MAX_PRICING, verifiedAt: "2026-08-04", status: "verified" },
   // Cursor Auto router — Cursor's published fixed token price (verified).
   { provider: "cursor", modelId: "auto", cost4: { input: 1.25, output: 6, cacheRead: 0.25, cacheWrite: 1.25 }, source: "https://docs.cursor.com/account/pricing + https://cursor.com/blog/aug-2025-pricing", verifiedAt: "2026-07-20", status: "verified" },
+  // Cursor-specific native models (verified from official docs).
+  { provider: "cursor", modelId: "grok-4.6", cost4: CURSOR_GROK_46, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "grok-4.5", cost4: CURSOR_GROK_45, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "composer-1", cost4: CURSOR_COMPOSER_1, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "composer-2.5", cost4: CURSOR_COMPOSER_25, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "composer-2.5-fast", cost4: CURSOR_COMPOSER_25_FAST, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "grok-4.6-fast", cost4: CURSOR_GROK_46_FAST, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "grok-4.5-fast", cost4: CURSOR_GROK_45_FAST, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "claude-opus-5-fast", cost4: CLAUDE_OPUS_FAST, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
+  { provider: "cursor", modelId: "claude-opus-4-8-fast", cost4: CLAUDE_OPUS_FAST, source: CURSOR_DOCS_PRICING, verifiedAt: "2026-08-26", status: "verified" },
 ];
 
 /**
@@ -195,6 +220,68 @@ export const VERIFIED_PRICE_OVERRIDES: readonly ExpectedPriceOverlay[] = [
     verifiedAt: "2026-08-18",
     status: "verified",
   },
+  // Command Code official model catalog rates (sourced from https://commandcode.ai/models.data).
+  { provider: "command-code", modelId: "zai-org/GLM-5.3", cost4: { input: 1.4, output: 4.4, cacheRead: 0.26, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "zai-org/glm-5.3", cost4: { input: 1.4, output: 4.4, cacheRead: 0.26, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified-derived" },
+  { provider: "command-code", modelId: "zai-org/GLM-5.2", cost4: { input: 1.4, output: 4.4, cacheRead: 0.26, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "zai-org/glm-5.2", cost4: { input: 1.4, output: 4.4, cacheRead: 0.26, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified-derived" },
+  { provider: "command-code", modelId: "zai-org/GLM-5.2-Fast", cost4: { input: 3, output: 10.25, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "zai-org/glm-5.2-fast", cost4: { input: 3, output: 10.25, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified-derived" },
+  { provider: "command-code", modelId: "zai-org/GLM-5.1", cost4: { input: 1.4, output: 4.4, cacheRead: 0.26, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "zai-org/glm-5.1", cost4: { input: 1.4, output: 4.4, cacheRead: 0.26, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified-derived" },
+  { provider: "command-code", modelId: "zai-org/GLM-5", cost4: { input: 1, output: 3.2, cacheRead: 0.2, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "zai-org/glm-5", cost4: { input: 1, output: 3.2, cacheRead: 0.2, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified-derived" },
+  { provider: "command-code", modelId: "meta/muse-spark-1.2-contributor", cost4: { input: 0.1, output: 0.2, cacheRead: 0.002, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "meta/muse-spark-1.2", cost4: { input: 1.25, output: 4.25, cacheRead: 0.15, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "meta/muse-spark-1.1", cost4: { input: 1.25, output: 4.25, cacheRead: 0.15, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "google/gemini-3.7-flash", cost4: { input: 0.75, output: 3.75, cacheRead: 0.075, cacheWrite: 0.04167 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "google/gemini-3.6-flash", cost4: { input: 1.5, output: 7.5, cacheRead: 0.15, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "google/gemini-3.5-flash", cost4: { input: 1.5, output: 9, cacheRead: 0.15, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "google/gemini-3.5-flash-lite", cost4: { input: 0.3, output: 2.5, cacheRead: 0.03, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "google/gemini-3.1-flash-lite", cost4: { input: 0.25, output: 1.5, cacheRead: 0.03, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "deepseek/deepseek-v4-pro", cost4: { input: 0.66, output: 1.98, cacheRead: 0.022, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "deepseek/deepseek-v4-flash", cost4: { input: 0.22, output: 0.66, cacheRead: 0.007, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "deepseek/deepseek-v4-flash-vision-exp", cost4: { input: 0.22, output: 0.66, cacheRead: 0.007, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.8-Max", cost4: { input: 2, output: 6, cacheRead: 0.25, cacheWrite: 2.5 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.8-27B", cost4: { input: 0.4, output: 3, cacheRead: 0.04, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.7-Max", cost4: { input: 2.5, output: 7.5, cacheRead: 0.5, cacheWrite: 3.13 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.7-Plus", cost4: { input: 0.4, output: 1.6, cacheRead: 0.08, cacheWrite: 0.5 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.7-Flash", cost4: { input: 0.03, output: 0.13, cacheRead: 0.006, cacheWrite: 0.038 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.6-Max-Preview", cost4: { input: 1.3, output: 7.8, cacheRead: 0.26, cacheWrite: 1.63 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "Qwen/Qwen3.6-Plus", cost4: { input: 0.5, output: 3, cacheRead: 0.1, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "MiniMaxAI/MiniMax-M3", cost4: { input: 0.3, output: 1.2, cacheRead: 0.06, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "MiniMaxAI/MiniMax-M2.7", cost4: { input: 0.3, output: 1.2, cacheRead: 0.06, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "MiniMaxAI/MiniMax-M2.5", cost4: { input: 0.3, output: 1.2, cacheRead: 0.03, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "stepfun/Step-3.7-Flash", cost4: { input: 0.2, output: 1.15, cacheRead: 0.04, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "stepfun/Step-3.5-Flash", cost4: { input: 0.1, output: 0.3, cacheRead: 0.02, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "xiaomi/mimo-v2.5-pro", cost4: { input: 0.435, output: 0.87, cacheRead: 0.0036, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "xiaomi/mimo-v2.5", cost4: { input: 0.14, output: 0.28, cacheRead: 0.0028, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "nvidia/nemotron-3-ultra-550b-a55b", cost4: { input: 0.6, output: 2.4, cacheRead: 0.12, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "tencent/hy3-paid", cost4: { input: 0.14, output: 0.58, cacheRead: 0.035, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "moonshotai/Kimi-K3", cost4: { input: 3, output: 15, cacheRead: 0.3, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "moonshotai/Kimi-K2.7-Code", cost4: { input: 0.95, output: 4, cacheRead: 0.19, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "moonshotai/Kimi-K2.7-Code-Highspeed", cost4: { input: 1.9, output: 8, cacheRead: 0.38, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "moonshotai/Kimi-K2.6", cost4: { input: 0.95, output: 4, cacheRead: 0.16, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "moonshotai/Kimi-K2.5", cost4: { input: 0.6, output: 3, cacheRead: 0.1, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-fable-5", cost4: { input: 10, output: 50, cacheRead: 1, cacheWrite: 12.5 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-opus-5", cost4: { input: 5, output: 25, cacheRead: 0.5, cacheWrite: 6.25 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-opus-4-8", cost4: { input: 5, output: 25, cacheRead: 0.5, cacheWrite: 6.25 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-opus-4-7", cost4: { input: 5, output: 25, cacheRead: 0.5, cacheWrite: 6.25 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-sonnet-5", cost4: { input: 2, output: 10, cacheRead: 0.2, cacheWrite: 2.5 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-sonnet-4-6", cost4: { input: 3, output: 15, cacheRead: 0.3, cacheWrite: 3.75 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "claude-haiku-4-5", cost4: { input: 1, output: 5, cacheRead: 0.1, cacheWrite: 1.25 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.6-sol", cost4: { input: 5, output: 30, cacheRead: 0.5, cacheWrite: 6.25 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.6-terra", cost4: { input: 2, output: 12, cacheRead: 0.2, cacheWrite: 2.5 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.6-luna", cost4: { input: 0.2, output: 1.2, cacheRead: 0.02, cacheWrite: 0.25 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.5", cost4: { input: 5, output: 30, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.4", cost4: { input: 2.5, output: 15, cacheRead: 0.25, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.4-mini", cost4: { input: 0.75, output: 4.5, cacheRead: 0.075, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "gpt-5.3-codex", cost4: { input: 2, output: 8, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "sakana/fugu-ultra", cost4: { input: 5, output: 30, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "xai/grok-4.6", cost4: { input: 2, output: 6, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "xai/grok-4.5", cost4: { input: 2, output: 6, cacheRead: 0.5, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "thinkingmachines/inkling", cost4: { input: 1, output: 4.05, cacheRead: 0.17, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
+  { provider: "command-code", modelId: "thinkingmachines/inkling-small", cost4: { input: 0.5, output: 1.2, cacheRead: 0.1, cacheWrite: 0 }, source: COMMAND_CODE_PRICING, verifiedAt: COMMAND_CODE_VERIFIED_AT, status: "verified" },
 ];
 
 export function findVerifiedPriceOverride(
@@ -223,6 +310,8 @@ export function findExpectedPriceOverlay(
 /** OpenAI Fast price multipliers retained as a compatibility export. */
 export const PRIORITY_MULTIPLIERS: Readonly<Record<string, number>> = {
   "gpt-5.6-sol": 2,
+  "gpt-daybreak-blue-latest": 2,
+  "daybreak-blue-latest": 2,
   // Post-price-cut Fast tables (https://openai.com/api-fast-mode/, 2026-08-05):
   // Terra 4/24/0.40/5 and Luna 0.40/2.40/0.04/0.50 are both 2× the corrected
   // standard tuples; the stale bases made these look like 1.6/0.4 (#907).
@@ -257,13 +346,19 @@ const XAI_PRIORITY_PRICING = "https://docs.x.ai/developers/advanced-api-usage/pr
  */
 export const PRIORITY_PRICING_RULES: readonly PriorityPricingRule[] = [
   ...["openai", "openai-apikey"].flatMap(provider =>
-    Object.entries(PRIORITY_MULTIPLIERS).map(([modelId, multiplier]): PriorityPricingRule => ({
-      provider,
-      modelId,
-      multiplier,
-      source: OPENAI_FAST_PRICING,
-      verifiedAt: "2026-08-05",
-    })),
+    Object.entries(PRIORITY_MULTIPLIERS)
+      // Daybreak's `gpt-` selector belongs to ChatGPT/Codex accounts; its bare selector belongs
+      // to the separately billed API-key catalog. All ordinary GPT ids remain valid on both.
+      .filter(([modelId]) => provider === "openai"
+        ? modelId !== "daybreak-blue-latest"
+        : modelId !== "gpt-daybreak-blue-latest")
+      .map(([modelId, multiplier]): PriorityPricingRule => ({
+        provider,
+        modelId,
+        multiplier,
+        source: OPENAI_FAST_PRICING,
+        verifiedAt: "2026-08-05",
+      })),
   ),
   ...["grok-4.5", "grok-4.6"].map((modelId): PriorityPricingRule => ({
     provider: "xai",
@@ -349,6 +444,29 @@ export const CONTEXT_TIERS: readonly ContextTier[] = [
       verifiedAt: "2026-08-03",
     })),
   ),
+  {
+    // The native Codex selector aliases gpt-5.6-sol and shares its 272k long-context tier.
+    provider: "openai",
+    modelId: "gpt-daybreak-blue-latest",
+    thresholdInputTokens: 272_000,
+    inclusive: false,
+    multiplier: OPENAI_LONG_CONTEXT,
+    confirmedPriorityRelation: "exclusive",
+    source: OPENAI_PRICING_DOC,
+    verifiedAt: "2026-08-11",
+  },
+  {
+    // The bare selector is the separately billed API-key alias. Daybreak Red has no tier row:
+    // the cyber snapshot's four long-context cells are all "-" on the pricing page.
+    provider: "openai-apikey",
+    modelId: "daybreak-blue-latest",
+    thresholdInputTokens: 272_000,
+    inclusive: false,
+    multiplier: OPENAI_LONG_CONTEXT,
+    confirmedPriorityRelation: "exclusive",
+    source: OPENAI_PRICING_DOC,
+    verifiedAt: "2026-08-11",
+  },
   {
     provider: "xai",
     modelId: "grok-4.5",
@@ -371,22 +489,6 @@ export const CONTEXT_TIERS: readonly ContextTier[] = [
     source: "https://docs.x.ai/developers/pricing",
     verifiedAt: "2026-08-18",
   },
-  {
-    // daybreak-blue-latest aliases gpt-5.6-sol, which publishes the full long-context row
-    // ($10 / $1 / $12.50 / $45). Scoped to openai-apikey ON PURPOSE: Daybreak is not
-    // routable on the Codex-login `openai` provider, so adding the alias to the shared
-    // OPENAI_GPT56_CONTEXT_MODELS list (expanded across both providers above) would mint a
-    // tier for a provider/model pair that cannot exist.
-    // daybreak-red-latest has NO tier row: the cyber snapshot's four long-context cells are
-    // all "-" on the pricing page (verified 2026-08-11).
-    provider: "openai-apikey",
-    modelId: "daybreak-blue-latest",
-    thresholdInputTokens: 272_000,
-    inclusive: false,
-    multiplier: OPENAI_LONG_CONTEXT,
-    source: OPENAI_PRICING_DOC,
-    verifiedAt: "2026-08-11",
-  },
   ...["minimax", "minimax-cn"].map((provider): ContextTier => ({
     provider,
     modelId: "MiniMax-M3",

package/src/usage/log.ts

@@ -13,6 +13,15 @@ import { CODEX_ACCOUNT_LOG_LABEL_RE } from "../codex/account-label";
 export type UsageStatus = "reported" | "unreported" | "unsupported" | "estimated";
 export type CodexUsageAccountLogLabel = "main" | `p${string}`;
 
+export const OAUTH_ACCOUNT_LOG_LABEL_RE = /^[a-f0-9]{8,64}(?:-\d+)?$/;
+
+export function isPersistableAccountLogLabel(value: unknown): value is string {
+  return (
+    value === "main"
+    || (typeof value === "string" && (CODEX_ACCOUNT_LOG_LABEL_RE.test(value) || OAUTH_ACCOUNT_LOG_LABEL_RE.test(value)))
+  );
+}
+
 export function isCodexUsageAccountLogLabel(value: unknown): value is CodexUsageAccountLogLabel {
   return value === "main" || (typeof value === "string" && CODEX_ACCOUNT_LOG_LABEL_RE.test(value));
 }
@@ -30,6 +39,7 @@ export type AttemptRecoveryKind =
   | "anthropic-oauth-429"
   | "cursor-oauth-auth"
   | "cursor-oauth-429"
+  | "oauth-account-429"
   | "image-413"
   | "opaque-blob-rejection"
   | "empty-completion";
@@ -52,8 +62,8 @@ export interface PersistedUsageAttempt {
   sendCount: number;
   recoveryKinds: AttemptRecoveryKind[];
   usageStatus: UsageStatus;
-  /** Stable non-PII identity for the Codex pool account that served this attempt. */
-  accountLogLabel?: CodexUsageAccountLogLabel;
+  /** Stable non-PII identity for the account that served this attempt. */
+  accountLogLabel?: string;
   inputTokenEstimate?: number;
   usage?: OcxUsage;
   totalTokens?: number;
@@ -70,6 +80,7 @@ export interface PersistedUsageAttempt {
 }
 
 export interface PersistedUsageEntry {
+  requestedAlias?: string;
   requestId: string;
   timestamp: number;
   provider: string;
@@ -81,8 +92,8 @@ export interface PersistedUsageEntry {
   admissionKind?: "configured" | "environment" | "loopback";
   /** The inbound wire, not the client product — see `surface`. */
   inboundProtocol?: "responses" | "chat" | "messages";
-  /** Stable non-PII identity for Codex Pool usage; absent for Direct/non-Codex traffic. */
-  accountLogLabel?: CodexUsageAccountLogLabel;
+  /** Stable non-PII identity for the account that served this request. */
+  accountLogLabel?: string;
   /** Best-effort chat/session correlation for Logs grouping (#330). */
   conversationId?: string;
   resolvedModel?: string;
@@ -220,8 +231,7 @@ const ATTEMPT_RECOVERY_KINDS = new Set<AttemptRecoveryKind>([
   "key-429",
   "rate-limit-429",
   "anthropic-oauth-429",
-  "cursor-oauth-auth",
-  "cursor-oauth-429",
+  "oauth-account-429",
   "image-413",
   "opaque-blob-rejection",
   "empty-completion",
@@ -367,7 +377,7 @@ function normalizeUsageAttempt(raw: unknown): PersistedUsageAttempt | null {
     sendCount: attempt.sendCount as number,
     recoveryKinds,
     usageStatus: attempt.usageStatus as UsageStatus,
-    ...(isCodexUsageAccountLogLabel(attempt.accountLogLabel)
+    ...(isPersistableAccountLogLabel(attempt.accountLogLabel)
       ? { accountLogLabel: attempt.accountLogLabel }
       : {}),
     ...(isNonNegativeFiniteNumber(attempt.inputTokenEstimate)
@@ -454,7 +464,7 @@ function normalizeUsageEntry(entry: PersistedUsageEntry): PersistedUsageEntry {
       : {}),
     ...(isKnownAdmissionKind(entry.admissionKind) ? { admissionKind: entry.admissionKind } : {}),
     ...(isKnownInboundProtocol(entry.inboundProtocol) ? { inboundProtocol: entry.inboundProtocol } : {}),
-    ...(isCodexUsageAccountLogLabel(entry.accountLogLabel)
+    ...(isPersistableAccountLogLabel(entry.accountLogLabel)
       ? { accountLogLabel: entry.accountLogLabel }
       : {}),
     ...(typeof entry.conversationId === "string" && entry.conversationId.trim()

package/src/usage/summary.ts

@@ -94,6 +94,7 @@ export interface UsageProvider {
 
 export interface UsageAccount {
   accountLogLabel: string;
+  provider?: string;
   ambiguous: boolean;
   requests: number;
   attemptCount: number;
@@ -684,12 +685,25 @@ function legacyCodexAccountLabel(provider: string): string | null {
   return suffix ?? LEGACY_AMBIGUOUS_ACCOUNT_LABEL;
 }
 
-function accountLabelForAttribution(provider: string, explicit: unknown): string | null {
-  if (isCodexUsageAccountLogLabel(explicit)) return explicit;
-  return legacyCodexAccountLabel(provider);
+export function accountLabelForAttribution(
+  provider: string,
+  explicit: unknown,
+  fallbackAccountByProvider?: Record<string, string>,
+): string | null {
+  if (typeof explicit === "string" && explicit.trim().length > 0) return explicit.trim();
+  const legacy = legacyCodexAccountLabel(provider);
+  if (legacy) return legacy;
+  const base = baseProviderLabel(provider);
+  if (fallbackAccountByProvider && fallbackAccountByProvider[base]) {
+    return fallbackAccountByProvider[base];
+  }
+  return null;
 }
 
-function buildAccounts(entries: PersistedUsageEntry[]): UsageAccount[] {
+function buildAccounts(
+  entries: PersistedUsageEntry[],
+  fallbackAccountByProvider?: Record<string, string>,
+): UsageAccount[] {
   const byLabel = new Map<string, UsageAccount>();
   const requestIds = new Map<string, Set<string>>();
 
@@ -702,12 +716,13 @@ function buildAccounts(entries: PersistedUsageEntry[]): UsageAccount[] {
     totalTokens?: number;
     estimate: ReturnType<typeof estimateRequestCost>;
   }): void => {
-    const label = accountLabelForAttribution(input.provider, input.accountLogLabel);
+    const label = accountLabelForAttribution(input.provider, input.accountLogLabel, fallbackAccountByProvider);
     if (!label) return;
     let row = byLabel.get(label);
     if (!row) {
       row = {
         accountLogLabel: label,
+        provider: baseProviderLabel(input.provider),
         ambiguous: label === LEGACY_AMBIGUOUS_ACCOUNT_LABEL,
         requests: 0,
         attemptCount: 0,
@@ -808,6 +823,7 @@ export function summarizeUsage(
   range: UsageRange,
   now: number,
   surface: UsageSurface = "all",
+  fallbackAccountByProvider?: Record<string, string>,
 ): UsageSummary {
   const { since } = rangeWindow(range, now);
   const filteredEntries = entries.filter(entry => {
@@ -837,7 +853,7 @@ export function summarizeUsage(
     days: buildDayGrid(range, since, now, filteredEntries),
     models: buildModels(filteredEntries, totals.totalTokens),
     providers: buildProviders(filteredEntries, totals.totalTokens),
-    accounts: buildAccounts(filteredEntries),
+    accounts: buildAccounts(filteredEntries, fallbackAccountByProvider),
   };
 }
 
@@ -869,12 +885,14 @@ function normalizeFilterValue(input: string | null | undefined): string | null {
  */
 export function projectUsageSummary<T extends UsageSummary>(
   summary: T,
-  filter: { provider?: string | null; model?: string | null },
+  filter: { provider?: string | null; model?: string | null; account?: string | null },
   entries?: PersistedUsageEntry[],
+  fallbackAccountByProvider?: Record<string, string>,
 ): T & { filter?: UsageFilterEcho } {
   const provider = normalizeFilterValue(filter.provider);
   const model = normalizeFilterValue(filter.model);
-  if (provider === null && model === null) return summary;
+  const account = normalizeFilterValue(filter.account);
+  if (provider === null && model === null && account === null) return summary;
 
   // Re-summarise from the entries the summary was built from, rather than
   // projecting over its rows.
@@ -891,9 +909,13 @@ export function projectUsageSummary<T extends UsageSummary>(
   //
   // The entries are already in hand on every path that filters, so the honest
   // computation is also the simple one.
-  const matches = (rowProvider: string, rowModel: string): boolean => {
+  const matches = (rowProvider: string, rowModel: string, rowAccount?: string | null): boolean => {
     if (provider !== null && baseProviderLabel(rowProvider).toLowerCase() !== provider) return false;
     if (model !== null && rowModel.toLowerCase() !== model) return false;
+    if (account !== null && rowAccount !== undefined) {
+      const effectiveAccount = accountLabelForAttribution(rowProvider, rowAccount, fallbackAccountByProvider);
+      if (!effectiveAccount || effectiveAccount.toLowerCase() !== account) return false;
+    }
     return true;
   };
 
@@ -909,10 +931,10 @@ export function projectUsageSummary<T extends UsageSummary>(
   const filtered: PersistedUsageEntry[] = [];
   for (const entry of source) {
     if (!entry.attempts?.length) {
-      if (matches(entry.provider, antigravityUsageModel(entry.provider, entry.model))) filtered.push(entry);
+      if (matches(entry.provider, antigravityUsageModel(entry.provider, entry.model), entry.accountLogLabel)) filtered.push(entry);
       continue;
     }
-    const attempts = entry.attempts.filter(a => matches(a.provider, antigravityUsageModel(a.provider, a.model)));
+    const attempts = entry.attempts.filter(a => matches(a.provider, antigravityUsageModel(a.provider, a.model), a.accountLogLabel ?? entry.accountLogLabel));
     if (attempts.length === 0) continue;
     // A combo is still counted once per participating model, so a filtered
     // request count can exceed the number of distinct requests. That is the
@@ -921,7 +943,7 @@ export function projectUsageSummary<T extends UsageSummary>(
     filtered.push({ ...entry, attempts });
   }
 
-  const projected = summarizeUsage(filtered, summary.range, summary.generatedAt, summary.surface);
+  const projected = summarizeUsage(filtered, summary.range, summary.generatedAt, summary.surface, fallbackAccountByProvider);
   // matched reflects usage inside the requested WINDOW, not anywhere in the
   // log: summarizeUsage applies the range and surface predicates, and the CLI
   // uses this flag to decide between a table and "no usage recorded".

package/src/web-search/exa-executor.ts

@@ -9,10 +9,11 @@
  * Never throws; every error string passes redactSecretString.
  */
 import { fetchWithResetRetry } from "../lib/upstream-retry";
-import { signalWithTimeout } from "../lib/abort";
+import { cancelBodyOnAbort, signalWithTimeout } from "../lib/abort";
+import { readBoundedResponseBytes } from "../lib/bounded-body";
 import { sidecarEnter } from "../lib/sidecar-tracker";
 import { redactSecretString } from "../lib/redact";
-import type { WebSearchSource } from "./parse";
+import { MAX_SIDECAR_RESPONSE_BYTES, type WebSearchSource } from "./parse";
 import type { SidecarOutcome, SidecarSettings } from "./executor";
 
 const EXA_SEARCH_URL = "https://api.exa.ai/search";
@@ -47,14 +48,44 @@ export async function runExaWebSearch(
       }),
       { abortSignal: linkedSignal.signal, label: "exa-web-search-sidecar" },
     );
-    if (!res.ok) {
-      const t = await res.text().catch(() => "");
-      // Scrub BEFORE truncating: slicing first can cut the literal key at the
-      // boundary, leaving an unscrubbable key prefix in the surviving text.
-      return { text: "", sources: [], error: `exa sidecar HTTP ${res.status}: ${scrub(t).slice(0, 200)}` };
+    const detachBodyGuard = cancelBodyOnAbort(res.body, linkedSignal.signal);
+    try {
+      let bounded: Awaited<ReturnType<typeof readBoundedResponseBytes>> | null = null;
+      try {
+        bounded = await readBoundedResponseBytes(res, {
+          maxBytes: MAX_SIDECAR_RESPONSE_BYTES,
+          signal: linkedSignal.signal,
+        });
+      } catch {
+        const reason = linkedSignal.signal.reason;
+        if (linkedSignal.signal.aborted && reason instanceof Error && reason.name === "TimeoutError") {
+          throw reason;
+        }
+        // Preserve the previous status-only/shapeless fallback for body read
+        // failures; fetch/header failures are still classified by the outer catch.
+      }
+      if (bounded?.oversized) {
+        const prefix = res.ok ? "exa sidecar response" : `exa sidecar HTTP ${res.status} response`;
+        return { text: "", sources: [], error: `${prefix} exceeded byte bound` };
+      }
+      // Response.text()/json() replace malformed UTF-8, so preserve that
+      // behavior while moving the byte bound ahead of decoding and parsing.
+      const text = bounded ? new TextDecoder().decode(bounded.bytes) : "";
+      if (!res.ok) {
+        // Scrub BEFORE truncating: slicing first can cut the literal key at the
+        // boundary, leaving an unscrubbable key prefix in the surviving text.
+        return { text: "", sources: [], error: `exa sidecar HTTP ${res.status}: ${scrub(text).slice(0, 200)}` };
+      }
+      let payload: unknown = null;
+      try {
+        payload = JSON.parse(text);
+      } catch {
+        // The mapper owns the stable malformed/empty JSON outcome.
+      }
+      return mapExaSearchResponse(payload);
+    } finally {
+      detachBodyGuard();
     }
-    const payload = await res.json().catch(() => null);
-    return mapExaSearchResponse(payload);
   } catch (e) {
     const kind = e instanceof Error && e.name === "TimeoutError" ? "timeout" : "connect_error";
     console.warn(`[web-search] exa sidecar ${kind} (${Date.now() - t0}ms)`);

package/src/web-search/index.ts

@@ -9,6 +9,7 @@ import { validateXaiSearchOptions, type XaiSearchOptions } from "./xai-executor"
 import type { OcxWebSearchSidecarConfig } from "../types";
 import { DEFAULT_STALL_TIMEOUT_SEC } from "../stall-timeout";
 import { buildWebSearchTool, extractHostedWebSearch, WEB_SEARCH_TOOL_NAME } from "./synthetic-tool";
+import { estimateInputTokens } from "../server/responses/input-admission";
 
 export { runWithWebSearch } from "./loop";
 export { buildWebSearchTool, extractHostedWebSearch, WEB_SEARCH_TOOL_NAME };
@@ -44,13 +45,19 @@ const STALL_MARGIN_SEC = 30;
  * deliberately permissive, so malformed values fall back locally without rejecting or rewriting
  * the caller's config object.
  */
-export function resolveRoutedModelStallTimeoutMs(value: unknown): number {
-  return typeof value === "number"
+export function resolveRoutedModelStallTimeoutMs(value: unknown, estimatedInputTokens?: number): number {
+  if (typeof value === "number"
     && Number.isInteger(value)
     && value >= 1
-    && value <= MAX_ROUTED_MODEL_STALL_TIMEOUT_MS
-    ? value
-    : DEFAULT_ROUTED_MODEL_STALL_TIMEOUT_MS;
+    && value <= MAX_ROUTED_MODEL_STALL_TIMEOUT_MS) {
+    return value;
+  }
+  const base = DEFAULT_ROUTED_MODEL_STALL_TIMEOUT_MS;
+  if (typeof estimatedInputTokens === "number" && Number.isFinite(estimatedInputTokens) && estimatedInputTokens > 100_000) {
+    const extraBlocks = Math.ceil((estimatedInputTokens - 100_000) / 100_000);
+    return Math.min(MAX_ROUTED_MODEL_STALL_TIMEOUT_MS, base + extraBlocks * 60_000);
+  }
+  return base;
 }
 
 function finiteCeil(value: number | undefined, fallback: number): number {
@@ -305,9 +312,10 @@ export function planWebSearch(
   const cfg = config.webSearchSidecar ?? {};
   if (cfg.enabled === false) return undefined;
   const timeoutMs = cfg.timeoutMs ?? DEFAULT_TIMEOUT_MS;
-  const routedModelStallTimeoutMs = resolveRoutedModelStallTimeoutMs(cfg.routedModelStallTimeoutMs);
-  // Same `?? 200_000` default the server applies when threading connectTimeoutMs into the loop.
-  const connectTimeoutMs = config.connectTimeoutMs ?? 200_000;
+  const safeModelId = typeof modelId === "string" ? modelId : "";
+  const estimatedInputTokens = estimateInputTokens(parsed, safeModelId);
+  const routedModelStallTimeoutMs = resolveRoutedModelStallTimeoutMs(cfg.routedModelStallTimeoutMs, estimatedInputTokens);
+  const connectTimeoutMs = config.connectTimeoutMs ?? Math.max(200_000, routedModelStallTimeoutMs);
   // Shared auth state (#2188): presence only — backend PREFERENCE stays with
   // resolveSidecarBackend's explicit-or-openai contract.
   const auth = resolveSidecarAuth(config);

package/src/web-search/loop.ts

@@ -318,10 +318,11 @@ export interface WebSearchLoopDeps {
   /** Called before each routed-model dispatch in the loop, for attempt telemetry. Same-target 429 replays pass the `rate-limit-429` recovery kind. */
   onAttemptSend?: (recovery?: AttemptRecoveryKind) => void;
   /**
-   * 429 key-failover hook: rotate the provider's active pool key and return a rebuilt adapter,
-   * or null when the pool is exhausted (same semantics as the normal routed path).
+   * 429 failover hook: rotate the provider's active credential and return a rebuilt adapter,
+   * or null when the pool is exhausted. Async hooks support OAuth refresh; existing synchronous
+   * key-pool hooks remain valid.
    */
-  on429?: (retryAfterHeader: string | null) => ProviderAdapter | null;
+  on429?: (retryAfterHeader: string | null) => ProviderAdapter | null | Promise<ProviderAdapter | null>;
   /** Opt-in same-target 429 policy (key-auth providers). When present, 429 replays on the SAME key before on429 rotation. */
   retryOn429Policy?: Required<RateLimitRetryPolicy> | null;
   /** Called only when the final bridged Responses stream reaches completed or incomplete. */
@@ -529,7 +530,7 @@ export async function runWithWebSearch(deps: WebSearchLoopDeps): Promise<Respons
       // 429 key-failover parity with the normal routed path: rotate pool keys until one responds
       // or the pool is exhausted (deps.on429 returns null — cooldown map guarantees termination).
       while (prepared.response.status === 429 && deps.on429) {
-        const rotated = deps.on429(prepared.response.headers.get("retry-after"));
+        const rotated = await deps.on429(prepared.response.headers.get("retry-after"));
         if (!rotated) break;
         // Never let a broken body's cancel promise outlive the cumulative header deadline. Observe
         // it, but proceed immediately to the rotated fetch under the SAME deadline signal.