4.1 maybe

src/info-page.ts

@@ -19,6 +19,9 @@ const MODEL_FAMILY_FRIENDLY_NAME: { [f in ModelFamily]: string } = {
   "gpt4-32k": "GPT-4 32k",
   "gpt4-turbo": "GPT-4 Turbo",
   gpt4o: "GPT-4o",
+  gpt41: "GPT-4.1",
+  "gpt41-mini": "GPT-4.1 Mini",
+  "gpt41-nano": "GPT-4.1 Nano",
   gpt45: "GPT-4.5",
   o1: "OpenAI o1",
   "o1-mini": "OpenAI o1 mini",
@@ -47,6 +50,9 @@ const MODEL_FAMILY_FRIENDLY_NAME: { [f in ModelFamily]: string } = {
   "azure-gpt4-turbo": "Azure GPT-4 Turbo",
   "azure-gpt4o": "Azure GPT-4o",
   "azure-gpt45": "Azure GPT-4.5",
+  "azure-gpt41": "Azure GPT-4.1",
+  "azure-gpt41-mini": "Azure GPT-4.1 Mini",
+  "azure-gpt41-nano": "Azure GPT-4.1 Nano",
   "azure-o1": "Azure o1",
   "azure-o1-mini": "Azure o1 mini",
   "azure-o3-mini": "Azure o3 mini",

src/proxy/middleware/request/preprocessors/validate-context-size.ts

@@ -60,6 +60,12 @@ export const validateContextSize: RequestPreprocessor = async (req) => {
     modelMax = 128000;
   } else if (model.match(/^gpt-4.5/)) {
     modelMax = 128000;
+  } else if (model.match(/^gpt-4\.1(-\d{4}-\d{2}-\d{2})?$/)) {
+    modelMax = 1000000;
+  } else if (model.match(/^gpt-4\.1-mini(-\d{4}-\d{2}-\d{2})?$/)) {
+    modelMax = 1000000;
+  } else if (model.match(/^gpt-4\.1-nano(-\d{4}-\d{2}-\d{2})?$/)) {
+    modelMax = 1000000;
   } else if (model.match(/^chatgpt-4o/)) {
     modelMax = 128000;
   } else if (model.match(/gpt-4-turbo(-\d{4}-\d{2}-\d{2})?$/)) {

src/shared/key-management/azure/provider.ts

@@ -74,6 +74,9 @@ export class AzureOpenAIKeyProvider implements KeyProvider<AzureOpenAIKey> {
         "azure-gpt4-turboTokens": 0,
         "azure-gpt4oTokens": 0,
         "azure-gpt45Tokens": 0,
+        "azure-gpt41Tokens": 0,
+        "azure-gpt41-miniTokens": 0,
+        "azure-gpt41-nanoTokens": 0,
         "azure-o1Tokens": 0,
         "azure-o1-miniTokens": 0,
         "azure-o3-miniTokens": 0,

src/shared/key-management/openai/provider.ts

@@ -114,6 +114,9 @@ export class OpenAIKeyProvider implements KeyProvider<OpenAIKey> {
         "gpt4-turboTokens": 0,
         gpt4oTokens: 0,
         gpt45Tokens: 0,
+        gpt41Tokens: 0,
+        "gpt41-minitokens": 0,
+        "gpt41-nanotokens": 0,
         "o1Tokens": 0,
         "o1-miniTokens": 0,
         "o3-miniTokens": 0,

src/shared/models.ts

@@ -24,6 +24,9 @@ export type OpenAIModelFamily =
   | "gpt4-32k"
   | "gpt4-turbo"
   | "gpt4o"
+  | "gpt41"
+  | "gpt41-mini"
+  | "gpt41-nano"
   | "gpt45"
   | "o1"
   | "o1-mini"
@@ -68,6 +71,9 @@ export const MODEL_FAMILIES = (<A extends readonly ModelFamily[]>(
   "gpt4-turbo",
   "gpt4o",
   "gpt45",
+  "gpt41",
+  "gpt41-mini",
+  "gpt41-nano",
   "o1",
   "o1-mini",
   "o3-mini",
@@ -95,6 +101,9 @@ export const MODEL_FAMILIES = (<A extends readonly ModelFamily[]>(
   "azure-gpt4-turbo",
   "azure-gpt4o",
   "azure-gpt45",
+  "azure-gpt41",
+  "azure-gpt41-mini",
+  "azure-gpt41-nano",
   "azure-dall-e",
   "azure-o1",
   "azure-o1-mini",
@@ -126,6 +135,9 @@ export const MODEL_FAMILY_SERVICE: {
   "gpt4-32k": "openai",
   gpt4o: "openai",
   gpt45: "openai",
+  gpt41: "openai",
+  "gpt41-mini": "openai",
+  "gpt41-nano": "openai",
   "o1": "openai",
   "o1-mini": "openai",
   "o3-mini": "openai",
@@ -146,6 +158,9 @@ export const MODEL_FAMILY_SERVICE: {
   "azure-gpt4-turbo": "azure",
   "azure-gpt4o": "azure",
   "azure-gpt45": "azure",
+  "azure-gpt41": "azure",
+  "azure-gpt41-mini": "azure",
+  "azure-gpt41-nano": "azure",
   "azure-dall-e": "azure",
   "azure-o1": "azure",
   "azure-o1-mini": "azure",
@@ -163,6 +178,9 @@ export const IMAGE_GEN_MODELS: ModelFamily[] = ["dall-e", "azure-dall-e"];
 
 export const OPENAI_MODEL_FAMILY_MAP: { [regex: string]: OpenAIModelFamily } = {
   "^gpt-4\\.5(-preview)?(-\\d{4}-\\d{2}-\\d{2})?$": "gpt45",
+  "^gpt-4\\.1(-\\d{4}-\\d{2}-\\d{2})?$": "gpt41",
+  "^gpt-4\\.1-mini(-\\d{4}-\\d{2}-\\d{2})?$": "gpt41-mini",
+  "^gpt-4\\.1-nano(-\\d{4}-\\d{2}-\\d{2})?$": "gpt41-nano",
   "^gpt-4o(-\\d{4}-\\d{2}-\\d{2})?$": "gpt4o",
   "^chatgpt-4o": "gpt4o",
   "^gpt-4o-mini(-\\d{4}-\\d{2}-\\d{2})?$": "turbo", // closest match

src/shared/stats.ts

@@ -6,6 +6,14 @@ import { ModelFamily } from "./models";
 export function getTokenCostUsd(model: ModelFamily, tokens: number) {
   let cost = 0;
   switch (model) {
+    case "gpt41":
+    case "azure-gpt41":
+    case "gpt41-mini":
+    case "azure-gpt41-mini":
+    case "gpt41-nano":
+    case "azure-gpt41-nano":
+      cost = 0.0000001;
+      break;
     case "gpt45":
     case "azure-gpt45":
       // $75/$150 for 1M input/output tokens pricing, averaged to $80