@tyvm/knowhow 0.0.159 → 0.0.161

package/src/clients/pricing/nvidia.ts

@@ -1,102 +0,0 @@
-/**
- * NVIDIA NIM pricing (USD per 1M tokens)
- * Source: https://build.nvidia.com/explore/discover
- * Many models are free (marked $0). Paid models use credit-based pricing.
- */
-export const NvidiaTextPricing: Record<string, { input: number; output: number }> = {
-  // Free models
-  "meta/llama-3.1-8b-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama-3.1-70b-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama-3.1-405b-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama-3.3-70b-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama-4-scout-17b-16e-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama-4-maverick-17b-128e-instruct": { input: 0.0, output: 0.0 },
-  "mistralai/mistral-7b-instruct-v0.3": { input: 0.0, output: 0.0 },
-  "mistralai/mixtral-8x7b-instruct-v0.1": { input: 0.0, output: 0.0 },
-  "mistralai/codestral-22b-instruct-v0.1": { input: 0.0, output: 0.0 },
-  "microsoft/phi-4-mini-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-4-multimodal-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3-medium-128k-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3-medium-4k-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3-small-128k-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3-small-8k-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3-vision-128k-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3.5-moe-instruct": { input: 0.0, output: 0.0 },
-  "microsoft/phi-3.5-vision-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen2.5-7b-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen2.5-72b-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen2.5-coder-32b-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen2.5-coder-7b-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen3-235b-a22b": { input: 0.0, output: 0.0 },
-  "qwen/qwen3-coder-480b-a35b-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen3-next-80b-a3b-instruct": { input: 0.0, output: 0.0 },
-  "qwen/qwen3-next-80b-a3b-thinking": { input: 0.0, output: 0.0 },
-  "qwen/qwen3.5-397b-a17b": { input: 0.0, output: 0.0 },
-  "qwen/qwq-32b": { input: 0.0, output: 0.0 },
-  "deepseek-ai/deepseek-r1": { input: 0.0, output: 0.0 },
-  "deepseek-ai/deepseek-r1-0528": { input: 0.0, output: 0.0 },
-  "deepseek-ai/deepseek-coder-6.7b-instruct": { input: 0.0, output: 0.0 },
-  "deepseek-ai/deepseek-v3.1": { input: 0.0, output: 0.0 },
-  "deepseek-ai/deepseek-v4-flash": { input: 0.14, output: 0.28 },
-  "deepseek-ai/deepseek-v3.1-terminus": { input: 0.0, output: 0.0 },
-  "deepseek-ai/deepseek-v3.2": { input: 0.0, output: 0.0 },
-  "nvidia/llama-3.1-nemotron-70b-instruct": { input: 0.0, output: 0.0 },
-  "nvidia/llama-3.1-nemotron-51b-instruct": { input: 0.0, output: 0.0 },
-  "nvidia/llama-3.1-nemotron-ultra-253b-v1": { input: 0.0, output: 0.0 },
-  "nvidia/llama-3.3-nemotron-super-49b-v1": { input: 0.0, output: 0.0 },
-  "nvidia/llama-3.3-nemotron-super-49b-v1.5": { input: 0.0, output: 0.0 },
-  "nvidia/llama-embed-nemotron-8b": { input: 0.0, output: 0.0 },
-  "nvidia/llama3-chatqa-1.5-70b": { input: 0.0, output: 0.0 },
-  "nvidia/cosmos-nemotron-34b": { input: 0.0, output: 0.0 },
-  "nvidia/nemoretriever-ocr-v1": { input: 0.0, output: 0.0 },
-  "nvidia/nemotron-3-nano-30b-a3b": { input: 0.0, output: 0.0 },
-  "nvidia/nvidia-nemotron-nano-9b-v2": { input: 0.0, output: 0.0 },
-  "nvidia/parakeet-tdt-0.6b-v2": { input: 0.0, output: 0.0 },
-  "google/gemma-3-27b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-3-12b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-3-4b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-3-1b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-3n-e2b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-3n-e4b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-4-31b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-2-27b-it": { input: 0.0, output: 0.0 },
-  "google/gemma-2-2b-it": { input: 0.0, output: 0.0 },
-  "google/codegemma-1.1-7b": { input: 0.0, output: 0.0 },
-  "google/codegemma-7b": { input: 0.0, output: 0.0 },
-  "meta/llama-3.2-11b-vision-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama-3.2-1b-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama3-70b-instruct": { input: 0.0, output: 0.0 },
-  "meta/llama3-8b-instruct": { input: 0.0, output: 0.0 },
-  "meta/codellama-70b": { input: 0.0, output: 0.0 },
-  "mistralai/mistral-large-2-instruct": { input: 0.0, output: 0.0 },
-  "mistralai/mistral-large-3-675b-instruct-2512": { input: 0.0, output: 0.0 },
-  "mistralai/mistral-small-3.1-24b-instruct-2503": { input: 0.0, output: 0.0 },
-  "mistralai/ministral-14b-instruct-2512": { input: 0.0, output: 0.0 },
-  "mistralai/mamba-codestral-7b-v0.1": { input: 0.0, output: 0.0 },
-  "mistralai/devstral-2-123b-instruct-2512": { input: 0.0, output: 0.0 },
-  "minimaxai/minimax-m2.1": { input: 0.0, output: 0.0 },
-  "minimaxai/minimax-m2.5": { input: 0.0, output: 0.0 },
-  "minimaxai/minimax-m2.7": { input: 0.0, output: 0.0 },
-  "moonshotai/kimi-k2-instruct": { input: 0.0, output: 0.0 },
-  "moonshotai/kimi-k2-instruct-0905": { input: 0.0, output: 0.0 },
-  "moonshotai/kimi-k2-thinking": { input: 0.0, output: 0.0 },
-  "moonshotai/kimi-k2.5": { input: 0.0, output: 0.0 },
-  "openai/gpt-oss-120b": { input: 0.0, output: 0.0 },
-  "openai/whisper-large-v3": { input: 0.0, output: 0.0 },
-  "stepfun-ai/step-3.5-flash": { input: 0.0, output: 0.0 },
-  "z-ai/glm-5.1": { input: 0.0, output: 0.0 },
-  "z-ai/glm4.7": { input: 0.0, output: 0.0 },
-  "z-ai/glm5": { input: 0.0, output: 0.0 },
-  // Nemotron 4 340B — free on NVIDIA NIM platform
-  // Source: models.dev (cost: $0), build.nvidia.com (free sign-up, no credit card)
-  "nvidia/nemotron-4-340b-instruct": { input: 0.0, output: 0.0 },
-  // Paid models
-  "deepseek-ai/deepseek-v4-pro": { input: 1.74, output: 3.48 },
-  "nvidia/nemotron-3-super-120b-a12b": { input: 0.2, output: 0.8 },
-};
-
-export const NvidiaImagePricing: Record<string, { input: number; output: number }> = {
-  "black-forest-labs/flux.1-dev": { input: 0.0, output: 0.0 },
-  "black-forest-labs/flux.1-schnell": { input: 0.0, output: 0.0 },
-  "stabilityai/stable-diffusion-3.5-medium": { input: 0.0, output: 0.0 },
-};

package/src/clients/pricing/openai.ts

@@ -1,409 +0,0 @@
-/**
- * OpenAI model IDs, pricing, and catalog.
- * Single source of truth for all OpenAI models.
- */
-import {
-  completions,
-  embeddings,
-  images,
-  videos,
-  audios,
-  transactions,
-  ModelCatalogEntry,
-  ModelPricing,
-} from "./types";
-
-// ─── Model IDs ────────────────────────────────────────────────────────────────
-
-export const OpenAiModels = {
-  GPT_56_Sol: "gpt-5.6-sol",
-  GPT_56_Terra: "gpt-5.6-terra",
-  GPT_56_Luna: "gpt-5.6-luna",
-  GPT_55: "gpt-5.5",
-  GPT_55_Pro: "gpt-5.5-pro",
-  GPT_54: "gpt-5.4",
-  GPT_54_Mini: "gpt-5.4-mini",
-  GPT_54_Nano: "gpt-5.4-nano",
-  GPT_54_Pro: "gpt-5.4-pro",
-  GPT_53_Chat: "gpt-5.3-chat-latest",
-  GPT_53_Codex: "gpt-5.3-codex",
-  GPT_53_Codex_Spark: "gpt-5.3-codex-spark",
-  GPT_52: "gpt-5.2",
-  GPT_52_Chat: "gpt-5.2-chat-latest",
-  GPT_52_Codex: "gpt-5.2-codex",
-  GPT_52_Pro: "gpt-5.2-pro",
-  GPT_51: "gpt-5.1",
-  GPT_51_Chat: "gpt-5.1-chat-latest",
-  GPT_51_Codex: "gpt-5.1-codex",
-  GPT_51_Codex_Max: "gpt-5.1-codex-max",
-  GPT_51_Codex_Mini: "gpt-5.1-codex-mini",
-  GPT_5: "gpt-5",
-  GPT_5_Pro: "gpt-5-pro",
-  GPT_5_Chat: "gpt-5-chat-latest",
-  GPT_5_Codex: "gpt-5-codex",
-  GPT_5_Mini: "gpt-5-mini",
-  GPT_5_Nano: "gpt-5-nano",
-  GPT_41: "gpt-4.1-2025-04-14",
-  GPT_41_Mini: "gpt-4.1-mini-2025-04-14",
-  GPT_41_Nano: "gpt-4.1-nano-2025-04-14",
-  GPT_45: "gpt-4.5-preview-2025-02-27",
-  GPT_4o: "gpt-4o-2024-08-06",
-  GPT_4o_Audio: "gpt-4o-audio-preview-2024-12-17",
-  GPT_4o_Realtime: "gpt-4o-realtime-preview-2024-12-17",
-  GPT_4o_Mini: "gpt-4o-mini-2024-07-18",
-  GPT_4o_Mini_Audio: "gpt-4o-mini-audio-preview-2024-12-17",
-  GPT_4o_Mini_Realtime: "gpt-4o-mini-realtime-preview-2024-12-17",
-  o1: "o1-2024-12-17",
-  o1_Pro: "o1-pro-2025-03-19",
-  o3: "o3-2025-04-16",
-  o3_Pro: "o3-pro-2025-01-31",
-  o4_Mini: "o4-mini-2025-04-16",
-  o3_Mini: "o3-mini-2025-01-31",
-  o1_Mini: "o1-mini-2024-09-12",
-  o1_Preview: "o1-preview",
-  o3_Deep_Research: "o3-deep-research",
-  o4_Mini_Deep_Research: "o4-mini-deep-research",
-  GPT_4o_Mini_Search: "gpt-4o-mini-search-preview-2025-03-11",
-  GPT_4o_Search: "gpt-4o-search-preview-2025-03-11",
-  GPT_4o_Transcribe: "gpt-4o-transcribe",
-  GPT_4o_Mini_Transcribe: "gpt-4o-mini-transcribe",
-  GPT_Realtime_15: "gpt-realtime-1.5",
-  GPT_Realtime_Mini: "gpt-realtime-mini",
-  GPT_Image_2: "gpt-image-2",
-  GPT_Image_15: "gpt-image-1.5",
-  GPT_Image_1: "gpt-image-1",
-  GPT_Image_1_Mini: "gpt-image-1-mini",
-  ChatGPT_Image: "chatgpt-image-latest",
-  TTS_1: "tts-1",
-  Whisper_1: "whisper-1",
-  DALL_E_3: "dall-e-3",
-  DALL_E_2: "dall-e-2",
-  Sora: "sora",
-  Sora_2: "sora-2",
-  Sora_2_Pro: "sora-2-pro",
-  GPT_4o_2024_05_13: "gpt-4o-2024-05-13",
-  GPT_4o_2024_11_20: "gpt-4o-2024-11-20",
-  GPT_35_Turbo: "gpt-3.5-turbo",
-  GPT_4: "gpt-4",
-  GPT_4_Turbo: "gpt-4-turbo",
-} as const;
-
-export const OpenAiEmbeddingModels: Record<string, string> = {
-  EmbeddingAda2: "text-embedding-ada-002",
-  EmbeddingLarge3: "text-embedding-3-large",
-  EmbeddingSmall3: "text-embedding-3-small",
-};
-
-// ─── Modality arrays ──────────────────────────────────────────────────────────
-
-export const OpenAiReasoningModels: string[] = [
-  OpenAiModels.GPT_5,
-  OpenAiModels.GPT_56_Sol,
-  OpenAiModels.GPT_56_Terra,
-  OpenAiModels.GPT_56_Luna,
-  OpenAiModels.GPT_5_Pro,
-  OpenAiModels.GPT_5_Mini,
-  OpenAiModels.GPT_5_Nano,
-  OpenAiModels.GPT_51,
-  OpenAiModels.GPT_51,
-  OpenAiModels.GPT_52,
-  OpenAiModels.GPT_52_Pro,
-  OpenAiModels.GPT_53_Chat,
-  OpenAiModels.GPT_53_Codex,
-  OpenAiModels.GPT_54,
-  OpenAiModels.GPT_54_Mini,
-  OpenAiModels.GPT_54_Nano,
-  OpenAiModels.GPT_54_Pro,
-  OpenAiModels.GPT_55,
-  OpenAiModels.GPT_55_Pro,
-  OpenAiModels.GPT_5_Mini,
-  OpenAiModels.GPT_5_Nano,
-  OpenAiModels.GPT_5_Pro,
-  OpenAiModels.o3,
-  OpenAiModels.o3_Mini,
-  OpenAiModels.o4_Mini,
-];
-
-export const OpenAiChatModels: string[] = [
-  OpenAiModels.GPT_4o,
-  OpenAiModels.GPT_4o_Mini,
-  OpenAiModels.GPT_4o_2024_11_20,
-  OpenAiModels.GPT_41,
-  OpenAiModels.GPT_41_Mini,
-  OpenAiModels.GPT_41_Nano,
-  OpenAiModels.GPT_56_Sol,
-  OpenAiModels.GPT_56_Terra,
-  OpenAiModels.GPT_56_Luna,
-  OpenAiModels.GPT_54,
-  OpenAiModels.GPT_54_Mini,
-  OpenAiModels.GPT_55,
-  OpenAiModels.GPT_54_Nano,
-  OpenAiModels.GPT_54_Pro,
-  OpenAiModels.GPT_53_Chat,
-  OpenAiModels.GPT_53_Codex,
-  OpenAiModels.GPT_51,
-  OpenAiModels.GPT_52,
-  OpenAiModels.o3,
-  OpenAiModels.o3_Mini,
-  OpenAiModels.o4_Mini,
-];
-
-export const OpenAiEmbeddingModelsList: string[] = [
-  OpenAiEmbeddingModels.EmbeddingAda2,
-  OpenAiEmbeddingModels.EmbeddingLarge3,
-  OpenAiEmbeddingModels.EmbeddingSmall3,
-];
-
-export const OpenAiResponsesOnlyModels: string[] = [
-  OpenAiModels.GPT_5,
-  OpenAiModels.GPT_5_Mini,
-  OpenAiModels.GPT_5_Nano,
-  OpenAiModels.GPT_52_Pro,
-  OpenAiModels.GPT_53_Codex,
-  OpenAiModels.GPT_54,
-  OpenAiModels.GPT_54_Mini,
-  OpenAiModels.GPT_54_Nano,
-  OpenAiModels.GPT_54_Pro,
-  OpenAiModels.GPT_55_Pro,
-  OpenAiModels.GPT_55,
-  OpenAiModels.GPT_5_Pro,
-  OpenAiModels.GPT_56_Sol,
-  OpenAiModels.GPT_56_Terra,
-  OpenAiModels.GPT_56_Luna,
-  OpenAiModels.o1,
-  OpenAiModels.o1_Pro,
-];
-
-// Models that exist in our catalog but have limited/no public access
-export const OpenAiLimitedAvailabilityModels: string[] = [
-  OpenAiModels.GPT_53_Codex_Spark, // 404 – not publicly available
-  OpenAiModels.o3_Pro, // 404 – not publicly available
-];
-
-// Image models
-export const OpenAiImageModels: string[] = [
-  OpenAiModels.DALL_E_3,
-  OpenAiModels.DALL_E_2,
-  OpenAiModels.GPT_Image_2,
-  OpenAiModels.GPT_Image_15,
-  OpenAiModels.GPT_Image_1_Mini,
-  OpenAiModels.ChatGPT_Image,
-];
-export const OpenAiDeprecatedImageModels: string[] = [OpenAiModels.GPT_Image_1];
-
-// Video models
-export const OpenAiVideoModels: string[] = [
-  OpenAiModels.Sora,
-  OpenAiModels.Sora_2,
-  OpenAiModels.Sora_2_Pro,
-];
-
-// Audio models
-export const OpenAiTTSModels: string[] = [OpenAiModels.TTS_1];
-export const OpenAiTranscriptionModels: string[] = [
-  OpenAiModels.Whisper_1,
-  OpenAiModels.GPT_4o_Transcribe,
-  OpenAiModels.GPT_4o_Mini_Transcribe,
-];
-export const OpenAiRealtimeModels: string[] = [
-  OpenAiModels.GPT_Realtime_15,
-  OpenAiModels.GPT_Realtime_Mini,
-];
-export const OpenAiAudioModels: string[] = [
-  ...OpenAiTTSModels,
-  ...OpenAiTranscriptionModels,
-  ...OpenAiRealtimeModels,
-];
-export const OpenAiDeprecatedAudioModels: string[] = [
-  OpenAiModels.GPT_4o_Audio,
-  OpenAiModels.GPT_4o_Mini_Audio,
-  OpenAiModels.GPT_4o_Realtime,
-  OpenAiModels.GPT_4o_Mini_Realtime,
-];
-
-// Search/transaction models
-export const OpenAiSearchModels: string[] = [
-  OpenAiModels.GPT_4o_Mini_Search,
-  OpenAiModels.GPT_4o_Search,
-];
-
-// ─── Pricing (USD per 1M tokens) ──────────────────────────────────────────────
-
-export const OpenAiTextPricing: Record<string, ModelPricing> = {
-  // gpt-5.6 series — short context / long context pricing
-  [OpenAiModels.GPT_56_Sol]: {
-    input: 4.0,
-    cached_input: 0.40,
-    cache_write: 5.00,
-    output: 30.0,
-    input_gt_200k: 10.0,
-    output_gt_200k: 30.0,
-    changelog: [
-      {
-        date: "2026-08-01",
-        note: "Launch pricing — promotional rates available through November 21, 2026",
-        pricing: { input: 5.0, cached_input: 0.5, cache_write: 6.25, output: 30.0, input_gt_200k: 10.0, output_gt_200k: 45.0 },
-      },
-    ],
-  },
-  [OpenAiModels.GPT_56_Terra]: { input: 2.0, cached_input: 0.20, cache_write: 2.5,   output: 12.0, input_gt_200k: 4.0,  output_gt_200k: 18.0 },
-  [OpenAiModels.GPT_56_Luna]:  { input: 0.2, cached_input: 0.02, cache_write: 0.25,  output: 1.2,  input_gt_200k: 0.4,  output_gt_200k: 1.8  },
-  [OpenAiModels.GPT_55]: { input: 5.0, cached_input: 0.5, output: 30.0 },
-  [OpenAiModels.GPT_55_Pro]: { input: 30.0, cached_input: 0, output: 180.0, reasoningLevels: ["medium", "high", "xhigh"] },
-  [OpenAiModels.GPT_54]: { input: 2.5, cached_input: 0.25, output: 15.0 },
-  [OpenAiModels.GPT_54_Mini]: { input: 0.75, cached_input: 0.075, output: 4.5 },
-  [OpenAiModels.GPT_54_Nano]: { input: 0.2, cached_input: 0.02, output: 1.25 },
-  [OpenAiModels.GPT_54_Pro]: { input: 30.0, cached_input: 0, output: 180.0 },
-  [OpenAiModels.GPT_53_Chat]: {
-    input: 1.75,
-    cached_input: 0.175,
-    output: 14.0,
-  },
-  [OpenAiModels.GPT_53_Codex]: {
-    input: 1.75,
-    cached_input: 0.175,
-    output: 14.0,
-  },
-  [OpenAiModels.GPT_53_Codex_Spark]: {
-    input: 1.75,
-    cached_input: 0.175,
-    output: 14.0,
-    limitedAvailability: true,
-  },
-  [OpenAiModels.GPT_52]: { input: 1.75, cached_input: 0.175, output: 14.0 },
-  [OpenAiModels.GPT_52_Chat]:  { input: 1.75, cached_input: 0.175, output: 14.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_52_Codex]: { input: 1.75, cached_input: 0.175, output: 14.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_52_Pro]: { input: 21.0, cached_input: 0, output: 168.0, reasoningLevels: ["medium", "high", "xhigh"] },
-  [OpenAiModels.GPT_51]: { input: 1.25, cached_input: 0.125, output: 10.0 },
-  [OpenAiModels.GPT_51_Chat]:      { input: 1.25, cached_input: 0.125, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_51_Codex]:     { input: 1.25, cached_input: 0.125, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_51_Codex_Max]: { input: 1.25, cached_input: 0.125, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_51_Codex_Mini]:{ input: 0.25, cached_input: 0.025, output: 2.0,  deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_5]: { input: 1.25, cached_input: 0.125, output: 10.0 },
-  [OpenAiModels.GPT_5_Pro]: { input: 15.0, cached_input: 0, output: 120.0, reasoningLevels: ["high"] },
-  [OpenAiModels.GPT_5_Chat]: { input: 1.25, cached_input: 0.125, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_5_Codex]: { input: 1.25, cached_input: 0.125, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_5_Mini]: { input: 0.25, cached_input: 0.025, output: 2.0 },
-  [OpenAiModels.GPT_5_Nano]: { input: 0.05, cached_input: 0.005, output: 0.4 },
-  [OpenAiModels.GPT_41]: { input: 2.0, cached_input: 0.5, output: 8.0 },
-  [OpenAiModels.GPT_41_Mini]: { input: 0.4, cached_input: 0.1, output: 1.6 },
-  [OpenAiModels.GPT_41_Nano]: { input: 0.1, cached_input: 0.025, output: 0.4, deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.GPT_45]: { input: 75.0, cached_input: 37.5, output: 150.0, deprecated: true, deprecationDate: "2025-07-14", replacedBy: OpenAiModels.GPT_41 },
-  [OpenAiModels.GPT_4o]: { input: 2.5, cached_input: 1.25, output: 10.0 },
-  [OpenAiModels.GPT_4o_Audio]:   { input: 2.5, cached_input: 0, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_4o_Realtime]: { input: 5.0, cached_input: 2.5, output: 20.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_4o_Mini]: { input: 0.15, cached_input: 0.075, output: 0.6 },
-  [OpenAiModels.GPT_4o_Mini_Audio]:   { input: 0.15, cached_input: 0, output: 0.6, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_4o_Mini_Realtime]: { input: 0.6, cached_input: 0.3, output: 2.4, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.o1]:     { input: 15.0, cached_input: 7.5, output: 60.0, deprecated: true, deprecationDate: "2026-10-23", replacedBy: OpenAiModels.o3 },
-  [OpenAiModels.o1_Pro]: { input: 150.0, cached_input: 0, output: 600.0, deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.o3]:     { input: 2.0, cached_input: 0.5, output: 8.0 },
-  [OpenAiModels.o3_Pro]: { input: 20.0, cached_input: 0, output: 80.0, limitedAvailability: true },
-  [OpenAiModels.o4_Mini]: { input: 1.1, cached_input: 0.275, output: 4.4 },
-  [OpenAiModels.o3_Mini]:    { input: 1.1, cached_input: 0.55, output: 4.4, deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.o1_Mini]:    { input: 1.1, cached_input: 0.55, output: 4.4, deprecated: true, deprecationDate: "2025-10-27", replacedBy: OpenAiModels.o4_Mini },
-  [OpenAiModels.o1_Preview]: { input: 15.0, cached_input: 7.5, output: 60.0, deprecated: true, deprecationDate: "2025-07-28", replacedBy: OpenAiModels.o3 },
-  [OpenAiModels.o3_Deep_Research]:      { input: 10.0, cached_input: 2.5, output: 40.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.o4_Mini_Deep_Research]: { input: 2.0,  cached_input: 0.5, output: 8.0,  deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_4o_Mini_Search]: { input: 0.15, cached_input: 0, output: 0.6,  deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_4o_Search]:      { input: 2.5,  cached_input: 0, output: 10.0, deprecated: true, deprecationDate: "2026-07-23" },
-  [OpenAiModels.GPT_4o_Transcribe]: {
-    input: 2.5,
-    cached_input: 0,
-    output: 10.0,
-  },
-  [OpenAiModels.GPT_4o_Mini_Transcribe]: {
-    input: 1.25,
-    cached_input: 0,
-    output: 5.0,
-  },
-  [OpenAiModels.GPT_Realtime_15]: {
-    input: 4.0,
-    cached_input: 0.4,
-    output: 16.0,
-  },
-  [OpenAiModels.GPT_Realtime_Mini]: {
-    input: 0.6,
-    cached_input: 0.06,
-    output: 2.4,
-  },
-  [OpenAiModels.GPT_Image_2]: { input: 8.0, cached_input: 2.0, output: 30.0 },
-  [OpenAiModels.GPT_Image_15]: { input: 5.0, cached_input: 1.25, output: 10.0 },
-  [OpenAiModels.GPT_Image_1]: { input: 5.0, cached_input: 1.25, output: 40.0, deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.GPT_Image_1_Mini]: { input: 2.0, cached_input: 0.2, output: 0 },
-  [OpenAiModels.ChatGPT_Image]: {
-    input: 5.0,
-    cached_input: 1.25,
-    output: 40.0,
-  },
-  [OpenAiModels.GPT_4o_2024_05_13]: { input: 5.0, cached_input: 0, output: 15.0, deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.GPT_4o_2024_11_20]: {
-    input: 2.5,
-    cached_input: 1.25,
-    output: 10.0,
-  },
-  [OpenAiModels.GPT_35_Turbo]: { input: 0.5, cached_input: 0, output: 1.5,  deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.GPT_4]:        { input: 30.0, cached_input: 0, output: 60.0, deprecated: true, deprecationDate: "2026-10-23" },
-  [OpenAiModels.GPT_4_Turbo]:  { input: 10.0, cached_input: 0, output: 30.0, deprecated: true, deprecationDate: "2026-10-23" },
-  // Embeddings
-  [OpenAiEmbeddingModels.EmbeddingAda2]: { input: 0.1, output: 0 },
-  [OpenAiEmbeddingModels.EmbeddingLarge3]: { input: 0.13, output: 0 },
-  [OpenAiEmbeddingModels.EmbeddingSmall3]: { input: 0.02, output: 0 },
-  // TTS / Whisper
-  [OpenAiModels.TTS_1]: { input: 15.0, output: 0 },
-  [OpenAiModels.Whisper_1]: { input: 0.006, output: 0 },
-  // Image generation (per-image)
-  [OpenAiModels.DALL_E_3]: { image_generation: 0.04 },
-  [OpenAiModels.DALL_E_2]: { image_generation: 0.02 },
-  // Video generation (per second)
-  [OpenAiModels.Sora]: { video_generation: 0.012 },
-  [OpenAiModels.Sora_2]: { video_generation: 0.015 },
-  [OpenAiModels.Sora_2_Pro]: { video_generation: 0.025 },
-};
-
-// ─── All completion model IDs (active + deprecated) ──────────────────────────
-// Deprecation/replacement metadata is embedded in OpenAiTextPricing entries.
-
-const OpenAiAllCompletionModels: string[] = [
-  ...OpenAiChatModels,
-  ...OpenAiResponsesOnlyModels,
-  ...OpenAiLimitedAvailabilityModels,
-  // Deprecated chat/codex variants
-  OpenAiModels.GPT_5_Chat,       OpenAiModels.GPT_5_Codex,
-  OpenAiModels.GPT_51_Chat,      OpenAiModels.GPT_51_Codex,
-  OpenAiModels.GPT_51_Codex_Max, OpenAiModels.GPT_51_Codex_Mini,
-  OpenAiModels.GPT_52_Chat,      OpenAiModels.GPT_52_Codex,
-  OpenAiModels.o3_Deep_Research, OpenAiModels.o4_Mini_Deep_Research,
-  // Deprecated reasoning
-  OpenAiModels.o3_Mini,  OpenAiModels.o4_Mini,
-  OpenAiModels.o1,       OpenAiModels.o1_Pro,
-  OpenAiModels.o1_Mini,  OpenAiModels.o1_Preview,
-  // Deprecated chat
-  OpenAiModels.GPT_41_Nano,
-  OpenAiModels.GPT_4o_2024_05_13,
-  OpenAiModels.GPT_45,
-  OpenAiModels.GPT_35_Turbo, OpenAiModels.GPT_4, OpenAiModels.GPT_4_Turbo,
-];
-
-const OpenAiAllImageModels: string[] = [
-  ...OpenAiImageModels,
-  ...OpenAiDeprecatedImageModels,
-];
-
-const OpenAiAllAudioModels: string[] = [
-  ...OpenAiAudioModels,
-  ...OpenAiDeprecatedAudioModels,
-];
-
-// ─── Catalog ──────────────────────────────────────────────────────────────────
-// Metadata (deprecated, deprecationDate, limitedAvailability, replacedBy) is
-// read directly from the OpenAiTextPricing entries — no need for separate groups.
-
-export const OPENAI_MODEL_CATALOG: ModelCatalogEntry[] = [
-  ...completions(OpenAiAllCompletionModels, "openai", OpenAiTextPricing),
-  ...embeddings(OpenAiEmbeddingModelsList,  "openai", OpenAiTextPricing),
-  ...images(OpenAiAllImageModels,           "openai", OpenAiTextPricing),
-  ...videos(OpenAiVideoModels,              "openai", OpenAiTextPricing),
-  ...audios(OpenAiAllAudioModels,           "openai", OpenAiTextPricing),
-  ...transactions(OpenAiSearchModels,       "openai", OpenAiTextPricing),
-];

package/src/clients/pricing/openrouter.ts

@@ -1,36 +0,0 @@
-/**
- * OpenRouter pricing (USD per 1M tokens)
- * Source: https://openrouter.ai/models
- * Models with `:free` suffix are free. Others are paid.
- * OpenRouter acts as an aggregator — prices may vary from original providers.
- */
-export const OpenRouterTextPricing: Record<string, { input: number; output: number }> = {
-  // Free models (`:free` suffix)
-  "deepseek/deepseek-r1:free": { input: 0.0, output: 0.0 },
-  "deepseek/deepseek-chat-v3-0324:free": { input: 0.0, output: 0.0 },
-  "meta-llama/llama-3.3-70b-instruct:free": { input: 0.0, output: 0.0 },
-  "meta-llama/llama-4-maverick:free": { input: 0.0, output: 0.0 },
-  "meta-llama/llama-4-scout:free": { input: 0.0, output: 0.0 },
-  "google/gemma-3-27b-it:free": { input: 0.0, output: 0.0 },
-  "google/gemma-3-12b-it:free": { input: 0.0, output: 0.0 },
-  "microsoft/phi-4:free": { input: 0.0, output: 0.0 },
-  "qwen/qwen3-235b-a22b:free": { input: 0.0, output: 0.0 },
-  "qwen/qwen3-30b-a3b:free": { input: 0.0, output: 0.0 },
-  "mistralai/mistral-7b-instruct:free": { input: 0.0, output: 0.0 },
-  "nousresearch/hermes-3-llama-3.1-405b:free": { input: 0.0, output: 0.0 },
-  // Paid models (popular)
-  "openai/gpt-4o": { input: 2.5, output: 10.0 },
-  "openai/gpt-4o-mini": { input: 0.15, output: 0.6 },
-  "anthropic/claude-3.5-sonnet": { input: 3.0, output: 15.0 },
-  "anthropic/claude-3-haiku": { input: 0.25, output: 1.25 },
-  // Source: openrouter.ai/api/v1/models (2026-04)
-  "deepseek/deepseek-r1": { input: 0.7, output: 2.5 },
-  "deepseek/deepseek-chat-v3-0324": { input: 0.2, output: 0.77 },
-  "meta-llama/llama-3.3-70b-instruct": { input: 0.12, output: 0.3 },
-  "google/gemini-2.0-flash-001": { input: 0.1, output: 0.4 },
-  // Source: openrouter.ai/api/v1/models (2026-04) — qwen3-235b-a22b = $0.455/$1.82
-  // Note: models.dev shows $0.15/$0.85 for dated variant, using OpenRouter live price
-  "qwen/qwen3-235b-a22b": { input: 0.455, output: 1.82 },
-  "qwen/qwen3-235b-a22b-07-25": { input: 0.15, output: 0.85 },
-  "mistralai/mistral-large-2411": { input: 2.0, output: 6.0 },
-};

package/src/clients/pricing/qwen.ts

@@ -1,80 +0,0 @@
-/**
- * QWEN Cloud pricing (USD per 1M tokens)
- * Source: https://home.qwencloud.com/data/api.json?product=AliyunDeliveryService&action=ListModelSeries
- * International DashScope endpoint: https://dashscope-intl.aliyuncs.com/compatible-mode/v1
- * Last updated: 2026-08-03
- *
- * Notes:
- * - Several Qwen models use tiered pricing based on input context length.
- *   We use the standard tier (≤256k) as the base price here.
- * - DeepSeek and GLM models are also available via QwenCloud's DashScope endpoint.
- * - Video/image/audio generation models have per-second/per-image pricing (not per-token)
- *   and are NOT included in QwenTextPricing.
- */
-
-export const QwenTextPricing: Record<string, { input: number; output: number; cached_input?: number }> = {
-
-  // ---------------------------------------------------------------------------
-  // Qwen3.8 series (newest flagship, released 2026-08)
-  // ---------------------------------------------------------------------------
-  // qwen3.8-max: $2/$6, implicit cache: $0.25, explicit cache creation: $2.5, explicit cache read: $0.17
-  "qwen3.8-max": { input: 2.00, output: 6.00, cached_input: 0.25 },
-
-  // ---------------------------------------------------------------------------
-  // Qwen3.7 series
-  // ---------------------------------------------------------------------------
-  // qwen3.7-plus: tiered — ≤256k: $0.4/$1.6, 256k-1M: $1.2/$4.8
-  // implicit cache: $0.08 (≤256k), explicit cache creation: $0.5, explicit cache read: $0.04
-  "qwen3.7-plus": { input: 0.40, output: 1.60, cached_input: 0.08 },
-  "qwen3.7-plus-2026-05-26": { input: 0.40, output: 1.60, cached_input: 0.08 },
-
-  // qwen3.7-flash: tiered — ≤32k: $0.03/$0.13, 32k-256k: $0.10/$0.40, 256k-1M: $0.20/$0.80
-  "qwen3.7-flash": { input: 0.03, output: 0.13, cached_input: 0.006 },
-  "qwen3.7-flash-2026-07-15": { input: 0.03, output: 0.13, cached_input: 0.006 },
-
-  // qwen3.7-max: $2.50/$7.50, cache: $0.50
-  "qwen3.7-max": { input: 2.50, output: 7.50, cached_input: 0.50 },
-  "qwen3.7-max-2026-06-08": { input: 2.50, output: 7.50, cached_input: 0.50 },
-  "qwen3.7-max-2026-05-20": { input: 2.50, output: 7.50 },
-
-  // ---------------------------------------------------------------------------
-  // Qwen3.6 series
-  // ---------------------------------------------------------------------------
-  // qwen3.6-flash: tiered — ≤256k: $0.25/$1.50, 256k-1M: $1.00/$4.00
-  "qwen3.6-flash": { input: 0.25, output: 1.50, cached_input: 0.025 },
-  "qwen3.6-flash-2026-04-16": { input: 0.25, output: 1.50 },
-  // qwen3.6-plus: tiered — ≤256k: $0.50/$3.00, 256k-1M: $2.00/$6.00
-  "qwen3.6-plus": { input: 0.50, output: 3.00, cached_input: 0.05 },
-  "qwen3.6-plus-2026-04-02": { input: 0.50, output: 3.00 },
-
-  // ---------------------------------------------------------------------------
-  // Legacy alias models (qwen-turbo, qwen-plus, qwen-max, qwen-long)
-  // These are rolling aliases that point to the latest generation.
-  // ---------------------------------------------------------------------------
-  "qwen-turbo": { input: 0.05, output: 0.20, cached_input: 0.005 },
-  "qwen-turbo-latest": { input: 0.05, output: 0.20, cached_input: 0.005 },
-  "qwen-plus": { input: 0.40, output: 1.20, cached_input: 0.04 },
-  "qwen-plus-latest": { input: 0.40, output: 1.20, cached_input: 0.04 },
-  "qwen-max": { input: 0.861, output: 3.44, cached_input: 0.086 },
-  "qwen-max-latest": { input: 0.861, output: 3.44, cached_input: 0.086 },
-  // qwen-long: extended context up to 10M tokens
-  "qwen-long": { input: 0.05, output: 0.20, cached_input: 0.005 },
-
-  // ---------------------------------------------------------------------------
-  // DeepSeek models (available via QwenCloud/DashScope endpoint)
-  // ---------------------------------------------------------------------------
-  // DeepSeek-V4-Flash: lightweight MoE, 284B params, 13B activated, 1M context
-  "deepseek-v4-flash": { input: 0.20, output: 0.40, cached_input: 0.04 },
-  "deepseek-v4-flash-0731": { input: 0.20, output: 0.40, cached_input: 0.04 },
-  // DeepSeek-V4-Pro: flagship MoE, 1.6T params, 49B activated, 1M context
-  "deepseek-v4-pro": { input: 2.40, output: 4.80, cached_input: 0.20 },
-  // DeepSeek-V3.2: sparse attention, 128k context
-  // explicit cache creation: $0.713, explicit cache read: $0.057
-  "deepseek-v3.2": { input: 0.57, output: 1.71, cached_input: 0.114 },
-
-  // ---------------------------------------------------------------------------
-  // GLM models (Zhipu AI, available via QwenCloud/DashScope endpoint)
-  // ---------------------------------------------------------------------------
-  // GLM-5.2-Fast-Preview: 1M context, high-speed inference
-  "glm-5.2-fast-preview": { input: 2.80, output: 8.80, cached_input: 0.56 },
-};