@kenkaiiii/gg-ai 5.48.0 → 5.49.1

package/dist/index.d.cts

@@ -3,6 +3,8 @@ import Anthropic from '@anthropic-ai/sdk';
 import OpenAI from 'openai';
 
 type Provider = "anthropic" | "xiaomi" | "openai" | "gemini" | "glm" | "moonshot" | "minimax" | "deepseek" | "openrouter" | "sakana" | "xai" | "palsu"
+/** Hugging Face Inference Providers router (OpenAI-compatible). */
+ | "huggingface"
 /** Locally hosted OpenAI-compatible server (Ollama, LM Studio, llama.cpp, vLLM). */
  | "local";
 type ThinkingLevel = "low" | "medium" | "high" | "xhigh" | "max" | "ultra";

package/dist/index.d.ts

@@ -3,6 +3,8 @@ import Anthropic from '@anthropic-ai/sdk';
 import OpenAI from 'openai';
 
 type Provider = "anthropic" | "xiaomi" | "openai" | "gemini" | "glm" | "moonshot" | "minimax" | "deepseek" | "openrouter" | "sakana" | "xai" | "palsu"
+/** Hugging Face Inference Providers router (OpenAI-compatible). */
+ | "huggingface"
 /** Locally hosted OpenAI-compatible server (Ollama, LM Studio, llama.cpp, vLLM). */
  | "local";
 type ThinkingLevel = "low" | "medium" | "high" | "xhigh" | "max" | "ultra";

package/dist/index.js

@@ -59,6 +59,7 @@ var PROVIDER_DISPLAY = {
   openrouter: "OpenRouter",
   sakana: "Sakana",
   xai: "xAI (Grok)",
+  huggingface: "Hugging Face",
   xiaomi: "Xiaomi (MiMo)",
   minimax: "MiniMax"
 };
@@ -2960,6 +2961,7 @@ var CODE_ASSIST_SUPPORTED_MODELS = /* @__PURE__ */ new Set([
   "gemini-3.5-flash",
   "gemini-3-flash",
   "gemini-3.1-flash-lite",
+  "gemini-3.7-flash",
   "gemini-2.5-pro",
   "gemini-2.5-flash",
   "gemma-4-31b-it",
@@ -2982,13 +2984,14 @@ var ACCOUNT_GATED_MODELS = /* @__PURE__ */ new Set([
   "gemini-3-flash",
   "gemini-3.5-flash",
   "gemini-3.1-pro-preview",
-  "gemini-3.1-pro-preview-customtools"
+  "gemini-3.1-pro-preview-customtools",
+  "gemini-3.7-flash"
 ]);
 function accountGatedMessage(model) {
   return `Your Google account isn't entitled to "${model}" over Gemini Code Assist OAuth, so the API reports it as not found. This is an account-access limit, not a ggcoder bug.`;
 }
 function accountGatedHint() {
-  return `Newer Gemini models (3.5 Flash, 3.1 Pro Preview) are available only to Code Assist Standard/Enterprise accounts with preview/GA access enabled by a cloud admin \u2014 free/personal accounts usually can't call them. Switch to Gemini 3.1 Flash Lite (it works on this account) with /model, or sign in with a Code Assist Standard/Enterprise account that has preview access.`;
+  return `Newer Gemini models (3.7 Flash, 3.5 Flash, 3.1 Pro Preview) are available only to Code Assist Standard/Enterprise accounts with preview/GA access enabled by a cloud admin \u2014 free/personal accounts usually can't call them. Switch to Gemini 3.1 Flash Lite (it works on this account) with /model, or sign in with a Code Assist Standard/Enterprise account that has preview access.`;
 }
 function formatErrorMessage(status, body, model) {
   if (status === 404 && !CODE_ASSIST_SUPPORTED_MODELS.has(model)) {
@@ -3666,6 +3669,18 @@ providerRegistry.register("openrouter", {
     baseUrl: options.baseUrl ?? "https://openrouter.ai/api/v1"
   })
 });
+providerRegistry.register("huggingface", {
+  // Hugging Face Inference Providers router — one HF token (hf.co/settings/tokens,
+  // "Make calls to Inference Providers" permission) routes to whichever hosted
+  // backend serves each open model. Chat Completions-compatible; model ids are
+  // Hub repo paths ("Qwen/Qwen3-Coder-480B-A35B-Instruct"), optionally with an
+  // ":auto"/":fastest"/":cheapest" provider-selection suffix. Billing follows
+  // each backend's per-token rates on the HF account (small free tier).
+  stream: (options) => streamOpenAI({
+    ...options,
+    baseUrl: options.baseUrl ?? "https://router.huggingface.co/v1"
+  })
+});
 providerRegistry.register("sakana", {
   // Sakana Fugu is a multi-agent system exposed as a standard LLM through the
   // OpenAI-compatible Sakana API. We ride the Chat Completions transport (the