Method fetchModels

app/src/utils/ModelManager.ts:261–289

* Fetch models from all sources (remote servers + refresh local cache)

()

   * Fetch models from all sources (remote servers + refresh local cache)
   */
  public async fetchModels(): Promise<ModelsResponse> {
    try {
      const allModels: Model[] = [];

      // Fetch from all inference addresses
      for (const address of this.inferenceAddresses) {
        const models = await this.fetchFromAddress(address);
        allModels.push(...models);
      }

      // Update remote models state
      this.remoteModels = allModels;

      // Refresh local model caches
      // llama.cpp models (Tauri only)
      if (isTauri()) {
        await NativeLlmManager.getInstance().refreshGgufCache();
      }
      // Note: GemmaModelManager doesn't have a refresh - it reads from localStorage synchronously

      this.notifyListeners();
      return this.listModels();
    } catch (error) {
      return {
        models: [],
        error: `Could not retrieve models: ${error instanceof Error ? error.message : String(error)}`
      };
    }
  }

  /**
   * Fetch models from a single address