@tyvm/knowhow 0.0.105 → 0.0.107

package/src/clients/pricing/models.ts

@@ -1,248 +1,19 @@
 /**
- * Model catalog: a single source of truth for all supported AI models,
- * their providers, types, display names, and pricing.
- *
- * Pricing is in USD per 1M tokens (or per-image / per-second for media models).
- * This is exported so the Knowhow backend (and other consumers) can import
- * model/pricing data without duplicating it.
+ * Assembles ALL_MODEL_CATALOG from per-provider catalogs.
+ * Each provider file is the single source of truth for its models.
  */
-
-import { Models, EmbeddingModels } from "../../types";
-import { OpenAiTextPricing } from "./openai";
-import { AnthropicTextPricing } from "./anthropic";
-import { GeminiPricing } from "./google";
-import { XaiTextPricing, XaiImagePricing, XaiVideoPricing } from "./xai";
 import { ModelPricing, ModelType, ModelCatalogEntry } from "./types";
+import { OPENAI_MODEL_CATALOG } from "./openai";
+import { ANTHROPIC_MODEL_CATALOG } from "./anthropic";
+import { GOOGLE_MODEL_CATALOG } from "./google";
+import { XAI_MODEL_CATALOG } from "./xai";
 
 export { ModelPricing, ModelType, ModelCatalogEntry };
 
-// ─── Platform markup ──────────────────────────────────────────────────────────
-
 /** 2.5% platform markup applied on top of all provider base rates */
 export const USAGE_MARKUP_PERCENT = 2.5 / 100;
 
-// ─── Helpers ──────────────────────────────────────────────────────────────────
-
-function completion(
-  id: string,
-  provider: string,
-  displayName: string,
-  pricingOverride?: Partial<ModelPricing>
-): ModelCatalogEntry {
-  const base =
-    (OpenAiTextPricing as Record<string, Partial<ModelPricing>>)[id] ||
-    (AnthropicTextPricing as Record<string, Partial<ModelPricing>>)[id] ||
-    (GeminiPricing as Record<string, Partial<ModelPricing>>)[id] ||
-    (XaiTextPricing as Record<string, Partial<ModelPricing>>)[id] ||
-    {};
-  return {
-    id,
-    provider,
-    type: "completion",
-    displayName,
-    pricing: {
-      input: 0,
-      output: 0,
-      ...base,
-      ...pricingOverride,
-    },
-  };
-}
-
-function embedding(
-  id: string,
-  provider: string,
-  displayName: string,
-  input: number
-): ModelCatalogEntry {
-  return {
-    id,
-    provider,
-    type: "embedding",
-    displayName,
-    pricing: { input, output: 0 },
-  };
-}
-
-function image(
-  id: string,
-  provider: string,
-  displayName: string,
-  pricing: Partial<ModelPricing>
-): ModelCatalogEntry {
-  return {
-    id,
-    provider,
-    type: "image",
-    displayName,
-    pricing: { input: 0, output: 0, ...pricing },
-  };
-}
-
-function video(
-  id: string,
-  provider: string,
-  displayName: string,
-  video_generation: number
-): ModelCatalogEntry {
-  return {
-    id,
-    provider,
-    type: "video",
-    displayName,
-    pricing: { input: 0, output: 0, video_generation },
-  };
-}
-
-function audio(
-  id: string,
-  provider: string,
-  displayName: string,
-  pricing: Partial<ModelPricing>
-): ModelCatalogEntry {
-  return {
-    id,
-    provider,
-    type: "audio",
-    displayName,
-    pricing: { input: 0, output: 0, ...pricing },
-  };
-}
-
-function transaction(
-  id: string,
-  provider: string,
-  displayName: string,
-  pricing: Partial<ModelPricing>
-): ModelCatalogEntry {
-  return {
-    id,
-    provider,
-    type: "transaction",
-    displayName,
-    pricing: { input: 0, output: 0, ...pricing },
-  };
-}
-
-// ─── OpenAI ───────────────────────────────────────────────────────────────────
-
-export const OPENAI_MODEL_CATALOG: ModelCatalogEntry[] = [
-  // Completion
-  completion(Models.openai.GPT_4o, "openai", "GPT-4o"),
-  completion(Models.openai.GPT_4o_Mini, "openai", "GPT-4o Mini"),
-  completion(Models.openai.GPT_41, "openai", "GPT-4.1"),
-  completion(Models.openai.GPT_41_Mini, "openai", "GPT-4.1 Mini"),
-  completion(Models.openai.GPT_41_Nano, "openai", "GPT-4.1 Nano"),
-  completion(Models.openai.GPT_45, "openai", "GPT-4.5 Preview"),
-  completion(Models.openai.o1, "openai", "o1"),
-  completion(Models.openai.o1_Mini, "openai", "o1 Mini"),
-  completion(Models.openai.o3, "openai", "o3"),
-  completion(Models.openai.o3_Mini, "openai", "o3 Mini"),
-  completion(Models.openai.o4_Mini, "openai", "o4 Mini"),
-  // Embedding
-  embedding(EmbeddingModels.openai.EmbeddingAda2, "openai", "Embedding Ada 002", OpenAiTextPricing[EmbeddingModels.openai.EmbeddingAda2]?.input ?? 0.1),
-  embedding(EmbeddingModels.openai.EmbeddingLarge3, "openai", "Embedding 3 Large", OpenAiTextPricing[EmbeddingModels.openai.EmbeddingLarge3]?.input ?? 0.13),
-  embedding(EmbeddingModels.openai.EmbeddingSmall3, "openai", "Embedding 3 Small", OpenAiTextPricing[EmbeddingModels.openai.EmbeddingSmall3]?.input ?? 0.02),
-  // Image generation
-  image(Models.openai.DALL_E_3, "openai", "DALL-E 3", { image_generation: 0.04 }),
-  image(Models.openai.DALL_E_2, "openai", "DALL-E 2", { image_generation: 0.02 }),
-  image(Models.openai.GPT_Image_15, "openai", "GPT Image 1.5", { input: OpenAiTextPricing[Models.openai.GPT_Image_15]?.input ?? 5.0, output: OpenAiTextPricing[Models.openai.GPT_Image_15]?.output ?? 10.0 }),
-  image(Models.openai.GPT_Image_1_Mini, "openai", "GPT Image 1 Mini", { input: OpenAiTextPricing[Models.openai.GPT_Image_1_Mini]?.input ?? 2.0 }),
-  // Video generation
-  video(Models.openai.Sora, "openai", "Sora", 0.012),
-  video(Models.openai.Sora_2, "openai", "Sora 2", 0.015),
-  // Audio
-  audio(Models.openai.TTS_1, "openai", "TTS-1", { input: 15.0 }),
-  audio(Models.openai.Whisper_1, "openai", "Whisper 1", { input: 0.006 }),
-  // Transaction / Search
-  transaction(Models.openai.GPT_4o_Mini_Search, "openai", "GPT-4o Mini Search", { input: OpenAiTextPricing[Models.openai.GPT_4o_Mini_Search]?.input ?? 0.15, output: OpenAiTextPricing[Models.openai.GPT_4o_Mini_Search]?.output ?? 0.6 }),
-  transaction(Models.openai.GPT_4o_Search, "openai", "GPT-4o Search", { input: OpenAiTextPricing[Models.openai.GPT_4o_Search]?.input ?? 2.5, output: OpenAiTextPricing[Models.openai.GPT_4o_Search]?.output ?? 10.0 }),
-];
-
-// ─── Anthropic ────────────────────────────────────────────────────────────────
-
-export const ANTHROPIC_MODEL_CATALOG: ModelCatalogEntry[] = [
-  completion(Models.anthropic.Opus4_5, "anthropic", "Claude Opus 4.5"),
-  completion(Models.anthropic.Sonnet4_5, "anthropic", "Claude Sonnet 4.5"),
-  completion(Models.anthropic.Opus4, "anthropic", "Claude Opus 4"),
-  completion(Models.anthropic.Sonnet4, "anthropic", "Claude Sonnet 4"),
-  completion(Models.anthropic.Haiku4_5, "anthropic", "Claude Haiku 4.5"),
-  completion(Models.anthropic.Sonnet3_7, "anthropic", "Claude Sonnet 3.7"),
-  completion(Models.anthropic.Sonnet3_5, "anthropic", "Claude Sonnet 3.5"),
-  completion(Models.anthropic.Haiku3, "anthropic", "Claude Haiku 3"),
-  completion(Models.anthropic.Opus3, "anthropic", "Claude Opus 3"),
-];
-
-// ─── Google ───────────────────────────────────────────────────────────────────
-
-export const GOOGLE_MODEL_CATALOG: ModelCatalogEntry[] = [
-  // Completion
-  completion(Models.google.Gemini_25_Pro, "google", "Gemini 2.5 Pro"),
-  completion(Models.google.Gemini_25_Flash, "google", "Gemini 2.5 Flash"),
-  completion(Models.google.Gemini_25_Flash_Lite, "google", "Gemini 2.5 Flash Lite"),
-  completion(Models.google.Gemini_20_Flash, "google", "Gemini 2.0 Flash"),
-  completion(Models.google.Gemini_15_Pro, "google", "Gemini 1.5 Pro"),
-  completion(Models.google.Gemini_15_Flash, "google", "Gemini 1.5 Flash"),
-  completion(Models.google.Gemini_15_Flash_8B, "google", "Gemini 1.5 Flash 8B"),
-  // Embedding
-  embedding(EmbeddingModels.google.Gemini_Embedding, "google", "Gemini Embedding", GeminiPricing[EmbeddingModels.google.Gemini_Embedding]?.input ?? 0),
-  // Image generation
-  image(Models.google.Gemini_20_Flash_Preview_Image_Generation, "google", "Gemini 2.0 Flash Image", {
-    input: GeminiPricing[Models.google.Gemini_20_Flash_Preview_Image_Generation]?.input ?? 0.1,
-    output: GeminiPricing[Models.google.Gemini_20_Flash_Preview_Image_Generation]?.output ?? 0.4,
-    image_generation: GeminiPricing[Models.google.Gemini_20_Flash_Preview_Image_Generation]?.image_generation ?? 0.039,
-  }),
-  image(Models.google.Gemini_25_Flash_Image, "google", "Gemini 2.5 Flash Image", {
-    input: GeminiPricing[Models.google.Gemini_25_Flash_Image]?.input ?? 0.3,
-    output: GeminiPricing[Models.google.Gemini_25_Flash_Image]?.output ?? 0.039,
-    image_generation: GeminiPricing[Models.google.Gemini_25_Flash_Image]?.image_generation ?? 0.039,
-  }),
-  image(Models.google.Gemini_31_Flash_Image_Preview, "google", "Gemini 3.1 Flash Image", {
-    input: GeminiPricing[Models.google.Gemini_31_Flash_Image_Preview]?.input ?? 0.5,
-    output: GeminiPricing[Models.google.Gemini_31_Flash_Image_Preview]?.output ?? 3.0,
-    image_generation: GeminiPricing[Models.google.Gemini_31_Flash_Image_Preview]?.image_generation ?? 0.045,
-  }),
-  image(Models.google.Gemini_3_Pro_Image_Preview, "google", "Gemini 3 Pro Image", {
-    input: GeminiPricing[Models.google.Gemini_3_Pro_Image_Preview]?.input ?? 2.0,
-    output: GeminiPricing[Models.google.Gemini_3_Pro_Image_Preview]?.output ?? 12.0,
-    image_generation: GeminiPricing[Models.google.Gemini_3_Pro_Image_Preview]?.image_generation ?? 0.134,
-  }),
-  image(Models.google.Imagen_3, "google", "Imagen 4", { image_generation: GeminiPricing[Models.google.Imagen_3]?.image_generation ?? 0.04 }),
-  image(Models.google.Imagen_4_Fast, "google", "Imagen 4 Fast", { image_generation: GeminiPricing[Models.google.Imagen_4_Fast]?.image_generation ?? 0.02 }),
-  image(Models.google.Imagen_4_Ultra, "google", "Imagen 4 Ultra", { image_generation: GeminiPricing[Models.google.Imagen_4_Ultra]?.image_generation ?? 0.06 }),
-  // Video generation
-  video(Models.google.Veo_2, "google", "Veo 2", GeminiPricing[Models.google.Veo_2]?.video_generation ?? 0.35),
-  video(Models.google.Veo_3, "google", "Veo 3", GeminiPricing[Models.google.Veo_3]?.video_generation ?? 0.4),
-  video(Models.google.Veo_3_Fast, "google", "Veo 3 Fast", GeminiPricing[Models.google.Veo_3_Fast]?.video_generation ?? 0.1),
-  // Audio (TTS)
-  audio(Models.google.Gemini_25_Flash_TTS, "google", "Gemini 2.5 Flash TTS", {
-    input: GeminiPricing[Models.google.Gemini_25_Flash_TTS]?.input ?? 0.5,
-    output_audio: GeminiPricing[Models.google.Gemini_25_Flash_TTS]?.output_audio ?? 10.0,
-    output: GeminiPricing[Models.google.Gemini_25_Flash_TTS]?.output_audio ?? 10.0,
-  }),
-  audio(Models.google.Gemini_25_Pro_TTS, "google", "Gemini 2.5 Pro TTS", {
-    input: GeminiPricing[Models.google.Gemini_25_Pro_TTS]?.input ?? 1.0,
-    output_audio: GeminiPricing[Models.google.Gemini_25_Pro_TTS]?.output_audio ?? 20.0,
-    output: GeminiPricing[Models.google.Gemini_25_Pro_TTS]?.output_audio ?? 20.0,
-  }),
-];
-
-// ─── xAI ──────────────────────────────────────────────────────────────────────
-
-export const XAI_MODEL_CATALOG: ModelCatalogEntry[] = [
-  completion(Models.xai.Grok4, "xai", "Grok 4"),
-  completion(Models.xai.Grok3Beta, "xai", "Grok 3 Beta"),
-  completion(Models.xai.Grok3MiniBeta, "xai", "Grok 3 Mini Beta"),
-  completion(Models.xai.Grok3FastBeta, "xai", "Grok 3 Fast Beta"),
-  completion(Models.xai.Grok21212, "xai", "Grok 2"),
-  // Image generation
-  image(Models.xai.GrokImagineImage, "xai", "Grok Imagine Image", { image_generation: XaiImagePricing["grok-imagine-image"]?.image_generation ?? 0.02 }),
-  image("grok-2-image-1212", "xai", "Grok 2 Image", { image_generation: XaiImagePricing["grok-2-image-1212"]?.image_generation ?? 0.07 }),
-  // Video generation
-  video(Models.xai.GrokImagineVideo, "xai", "Grok Imagine Video", XaiVideoPricing["grok-imagine-video"]?.video_generation ?? 0.05),
-];
-
-// ─── Combined catalog ─────────────────────────────────────────────────────────
+export { OPENAI_MODEL_CATALOG, ANTHROPIC_MODEL_CATALOG, GOOGLE_MODEL_CATALOG, XAI_MODEL_CATALOG };
 
 export const ALL_MODEL_CATALOG: ModelCatalogEntry[] = [
   ...OPENAI_MODEL_CATALOG,

package/src/clients/pricing/nvidia.ts

@@ -0,0 +1,102 @@
+/**
+ * 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 },
+};