@juspay/neurolink 8.23.2 → 8.25.0

package/CHANGELOG.md

@@ -1,3 +1,15 @@
+## [8.25.0](https://github.com/juspay/neurolink/compare/v8.24.0...v8.25.0) (2025-12-30)
+
+### Features
+
+- **(observability):** Add support for custom metadata in Context ([b175249](https://github.com/juspay/neurolink/commit/b175249c61357b0e6d127932bd7824d0bfe6f2ed))
+
+## [8.24.0](https://github.com/juspay/neurolink/compare/v8.23.2...v8.24.0) (2025-12-28)
+
+### Features
+
+- **(openrouter):** add OpenRouter provider with 300+ model support ([563611f](https://github.com/juspay/neurolink/commit/563611f84c154e2966aebb6e8a414fcb60a26fd3)), closes [#608](https://github.com/juspay/neurolink/issues/608)
+
 ## [8.23.2](https://github.com/juspay/neurolink/compare/v8.23.1...v8.23.2) (2025-12-27)
 
 ### Bug Fixes

package/README.md

@@ -11,9 +11,9 @@ Enterprise AI development platform with unified provider access, production-read
 
 ## 🧠 What is NeuroLink?
 
-**NeuroLink is the universal AI integration platform that unifies 12 major AI providers and 100+ models under one consistent API.**
+**NeuroLink is the universal AI integration platform that unifies 13 major AI providers and 100+ models under one consistent API.**
 
-Extracted from production systems at Juspay and battle-tested at enterprise scale, NeuroLink provides a production-ready solution for integrating AI into any application. Whether you're building with OpenAI, Anthropic, Google, AWS Bedrock, Azure, or any of our 12 supported providers, NeuroLink gives you a single, consistent interface that works everywhere.
+Extracted from production systems at Juspay and battle-tested at enterprise scale, NeuroLink provides a production-ready solution for integrating AI into any application. Whether you're building with OpenAI, Anthropic, Google, AWS Bedrock, Azure, or any of our 13 supported providers, NeuroLink gives you a single, consistent interface that works everywhere.
 
 **Why NeuroLink?** Switch providers with a single parameter change, leverage 64+ built-in tools and MCP servers, deploy with confidence using enterprise features like Redis memory and multi-provider failover, and optimize costs automatically with intelligent routing. Use it via our professional CLI or TypeScript SDK—whichever fits your workflow.
 
@@ -30,6 +30,7 @@ Extracted from production systems at Juspay and battle-tested at enterprise scal
 - **PDF File Support** – Process PDF documents with native visual analysis for Vertex AI, Anthropic, Bedrock, AI Studio. → [PDF Guide](docs/features/pdf-support.md)
 - **LiteLLM Integration** – Access 100+ AI models from all major providers through unified interface. → [Setup Guide](docs/LITELLM-INTEGRATION.md)
 - **SageMaker Integration** – Deploy and use custom trained models on AWS infrastructure. → [Setup Guide](docs/SAGEMAKER-INTEGRATION.md)
+- **OpenRouter Integration** – Access 300+ models from OpenAI, Anthropic, Google, Meta, and more through a single unified API. → [Setup Guide](docs/getting-started/providers/openrouter.md)
 - **Human-in-the-loop workflows** – Pause generation for user approval/input before tool execution. → [HITL Guide](docs/features/hitl.md)
 - **Guardrails middleware** – Block PII, profanity, and unsafe content with built-in filtering. → [Guardrails Guide](docs/features/guardrails.md)
 - **Context summarization** – Automatic conversation compression for long-running sessions. → [Summarization Guide](docs/CONTEXT-SUMMARIZATION.md)
@@ -55,7 +56,7 @@ NeuroLink is a comprehensive AI development platform. Every feature below is pro
 
 ### 🤖 AI Provider Integration
 
-**12 providers unified under one API** - Switch providers with a single parameter change.
+**13 providers unified under one API** - Switch providers with a single parameter change.
 
 | Provider              | Models                         | Free Tier       | Tool Support | Status        | Documentation                                                           |
 | --------------------- | ------------------------------ | --------------- | ------------ | ------------- | ----------------------------------------------------------------------- |
@@ -66,6 +67,7 @@ NeuroLink is a comprehensive AI development platform. Every feature below is pro
 | **Google Vertex**     | Gemini via GCP                 | ❌              | ✅ Full      | ✅ Production | [Setup Guide](docs/getting-started/provider-setup.md#vertex)            |
 | **Azure OpenAI**      | GPT-4, GPT-4o, o1              | ❌              | ✅ Full      | ✅ Production | [Setup Guide](docs/getting-started/provider-setup.md#azure)             |
 | **LiteLLM**           | 100+ models unified            | Varies          | ✅ Full      | ✅ Production | [Setup Guide](docs/LITELLM-INTEGRATION.md)                              |
+| **OpenRouter**        | 300+ models unified            | Varies          | ✅ Full      | ✅ Production | [Setup Guide](docs/getting-started/provider-setup.md#openrouter)        |
 | **AWS SageMaker**     | Custom deployed models         | ❌              | ✅ Full      | ✅ Production | [Setup Guide](docs/SAGEMAKER-INTEGRATION.md)                            |
 | **Mistral AI**        | Mistral Large, Small           | ✅ Free Tier    | ✅ Full      | ✅ Production | [Setup Guide](docs/getting-started/provider-setup.md#mistral)           |
 | **Hugging Face**      | 100,000+ models                | ✅ Free         | ⚠️ Partial   | ✅ Production | [Setup Guide](docs/getting-started/provider-setup.md#huggingface)       |
@@ -297,7 +299,7 @@ Full command and API breakdown lives in [`docs/cli/commands.md`](docs/cli/comman
 
 | Capability               | Highlights                                                                                                               |
 | ------------------------ | ------------------------------------------------------------------------------------------------------------------------ |
-| **Provider unification** | 12+ providers with automatic fallback, cost-aware routing, provider orchestration (Q3).                                  |
+| **Provider unification** | 13+ providers with automatic fallback, cost-aware routing, provider orchestration (Q3).                                  |
 | **Multimodal pipeline**  | Stream images + CSV data + PDF documents across providers with local/remote assets. Auto-detection for mixed file types. |
 | **Quality & governance** | Auto-evaluation engine (Q3), guardrails middleware (Q4), HITL workflows (Q4), audit logging.                             |
 | **Memory & context**     | Conversation memory, Mem0 integration, Redis history export (Q4), context summarization (Q4).                            |

package/dist/adapters/providerImageAdapter.js

@@ -37,6 +37,7 @@ const IMAGE_LIMITS = {
     mistral: 10, // Conservative limit for Mistral
     // Note: Bedrock limit defined for future use when vision support is added
     bedrock: 20, // Same as Anthropic for Claude models on Bedrock
+    openrouter: 10, // Conservative limit, routes to various underlying providers
 };
 /**
  * Vision capability definitions for each provider
@@ -231,6 +232,37 @@ const VISION_CAPABILITIES = {
         // Groq models via LiteLLM (vision)
         "groq/llama-3.2-11b-vision-preview",
     ],
+    openrouter: [
+        // OpenRouter provides access to vision-capable models from multiple providers
+        // Anthropic Claude models (via OpenRouter)
+        "anthropic/claude-3-5-sonnet",
+        "anthropic/claude-3-5-haiku",
+        "anthropic/claude-3-opus",
+        "anthropic/claude-3-sonnet",
+        "anthropic/claude-3-haiku",
+        // OpenAI models (via OpenRouter)
+        "openai/gpt-4o",
+        "openai/gpt-4o-mini",
+        "openai/gpt-4-turbo",
+        "openai/gpt-4-vision-preview",
+        // Google models (via OpenRouter)
+        "google/gemini-2.5-pro",
+        "google/gemini-2.5-flash",
+        "google/gemini-2.0-flash",
+        "google/gemini-2.0-flash-001",
+        "google/gemini-1.5-pro",
+        "google/gemini-1.5-flash",
+        "google/gemini-pro-vision",
+        // Meta Llama models (vision-capable via OpenRouter)
+        "meta-llama/llama-3.2-90b-vision-instruct",
+        "meta-llama/llama-3.2-11b-vision-instruct",
+        // Pixtral/Mistral models (via OpenRouter)
+        "mistralai/pixtral-12b",
+        "mistralai/pixtral-large",
+        // Qwen models (via OpenRouter)
+        "qwen/qwen-2-vl-72b-instruct",
+        "qwen/qwen-2-vl-7b-instruct",
+    ],
     mistral: [
         // Mistral Large (latest has vision via Pixtral integration)
         "mistral-large-latest",
@@ -430,6 +462,11 @@ export class ProviderImageAdapter {
                     this.validateImageCount(images.length, "bedrock");
                     adaptedPayload = this.formatForAnthropic(text, images, true);
                     break;
+                case "openrouter":
+                    // OpenRouter routes to underlying providers, use OpenAI format
+                    this.validateImageCount(images.length, "openrouter");
+                    adaptedPayload = this.formatForOpenAI(text, images);
+                    break;
                 default:
                     throw new Error(`Vision not supported for provider: ${provider}`);
             }

package/dist/cli/commands/setup.js

@@ -125,6 +125,19 @@ const PROVIDERS = [
         pricing: "Free tier → €7 per 1M tokens",
         setupCommand: "neurolink setup --provider mistral",
     },
+    {
+        id: "openrouter",
+        name: "OpenRouter",
+        emoji: "🔀",
+        description: "100+ models, unified API access",
+        setupTime: "2 min",
+        cost: "Pay-per-use",
+        bestFor: "Model variety",
+        models: "GPT-4, Claude, Gemini, Llama, Mixtral, etc.",
+        strengths: "Access 100+ models via single API, model routing, fallbacks",
+        pricing: "Per-model pricing, free tier available",
+        setupCommand: "neurolink setup --provider openrouter",
+    },
 ];
 /**
  * Main setup command handler
@@ -270,6 +283,9 @@ async function checkExistingConfigurations() {
     if (process.env.MISTRAL_API_KEY) {
         configured.push("mistral");
     }
+    if (process.env.OPENROUTER_API_KEY) {
+        configured.push("openrouter");
+    }
     return configured;
 }
 /**
@@ -381,12 +397,46 @@ async function delegateToProviderSetup(providerId) {
         case "mistral":
             await handleMistralSetup(setupArgs);
             break;
+        case "openrouter":
+            await handleOpenRouterSetup();
+            break;
         default:
             throw new Error(`Unknown provider: ${providerId}`);
     }
     // After successful setup, show completion message
     await showSetupCompletion(providerId);
 }
+/**
+ * Handle OpenRouter provider setup
+ */
+async function handleOpenRouterSetup() {
+    logger.always("");
+    logger.always(chalk.blue("🔀 OpenRouter Setup"));
+    logger.always("");
+    logger.always("OpenRouter provides unified access to 100+ AI models from multiple providers");
+    logger.always("including OpenAI, Anthropic, Google, Meta, Mistral, and more.");
+    logger.always("");
+    logger.always(chalk.yellow("Step 1: Get your API key"));
+    logger.always("  1. Go to https://openrouter.ai/keys");
+    logger.always("  2. Sign in or create an account");
+    logger.always("  3. Create a new API key");
+    logger.always("");
+    logger.always(chalk.yellow("Step 2: Set the environment variable"));
+    logger.always("  Add to your shell profile (~/.bashrc, ~/.zshrc, etc.):");
+    logger.always("");
+    logger.always(chalk.cyan("  export OPENROUTER_API_KEY=your_api_key_here"));
+    logger.always("");
+    logger.always(chalk.yellow("Step 3: Test the configuration"));
+    logger.always(chalk.cyan('  neurolink generate "Hello!" --provider openrouter --model google/gemini-2.0-flash-exp:free'));
+    logger.always("");
+    logger.always(chalk.green("Available models include:"));
+    logger.always("  • anthropic/claude-3.5-sonnet - Best for analysis");
+    logger.always("  • openai/gpt-4o - Industry standard");
+    logger.always("  • google/gemini-2.0-flash-exp:free - Free tier");
+    logger.always("  • meta-llama/llama-3.1-70b-instruct - Open source");
+    logger.always("");
+    logger.always(chalk.gray("See all models at: https://openrouter.ai/models"));
+}
 /**
  * Show setup completion message
  */

package/dist/cli/factories/commandFactory.js

@@ -29,6 +29,8 @@ export class CLICommandFactory {
                 "auto",
                 "openai",
                 "openai-compatible",
+                "openrouter",
+                "or",
                 "bedrock",
                 "vertex",
                 "googleVertex",
@@ -797,6 +799,7 @@ export class CLICommandFactory {
                     choices: [
                         "google-ai",
                         "openai",
+                        "openrouter",
                         "anthropic",
                         "azure",
                         "bedrock",

package/dist/constants/enums.d.ts

@@ -5,6 +5,7 @@ export declare enum AIProviderName {
     BEDROCK = "bedrock",
     OPENAI = "openai",
     OPENAI_COMPATIBLE = "openai-compatible",
+    OPENROUTER = "openrouter",
     VERTEX = "vertex",
     ANTHROPIC = "anthropic",
     AZURE = "azure",
@@ -16,6 +17,24 @@ export declare enum AIProviderName {
     SAGEMAKER = "sagemaker",
     AUTO = "auto"
 }
+/**
+ * Popular Models for OpenRouter (300+ available at openrouter.ai/models)
+ * OpenRouter uses 'provider/model' format
+ */
+export declare enum OpenRouterModels {
+    CLAUDE_3_5_SONNET = "anthropic/claude-3-5-sonnet",
+    CLAUDE_3_5_HAIKU = "anthropic/claude-3-5-haiku",
+    CLAUDE_3_OPUS = "anthropic/claude-3-opus",
+    GPT_4O = "openai/gpt-4o",
+    GPT_4O_MINI = "openai/gpt-4o-mini",
+    GPT_4_TURBO = "openai/gpt-4-turbo",
+    GEMINI_2_0_FLASH = "google/gemini-2.0-flash",
+    GEMINI_1_5_PRO = "google/gemini-1.5-pro",
+    LLAMA_3_1_70B = "meta-llama/llama-3.1-70b-instruct",
+    LLAMA_3_1_8B = "meta-llama/llama-3.1-8b-instruct",
+    MISTRAL_LARGE = "mistralai/mistral-large",
+    MIXTRAL_8X7B = "mistralai/mixtral-8x7b-instruct"
+}
 /**
  * Supported Models for Amazon Bedrock
  */

package/dist/constants/enums.js

@@ -9,6 +9,7 @@ export var AIProviderName;
     AIProviderName["BEDROCK"] = "bedrock";
     AIProviderName["OPENAI"] = "openai";
     AIProviderName["OPENAI_COMPATIBLE"] = "openai-compatible";
+    AIProviderName["OPENROUTER"] = "openrouter";
     AIProviderName["VERTEX"] = "vertex";
     AIProviderName["ANTHROPIC"] = "anthropic";
     AIProviderName["AZURE"] = "azure";
@@ -20,6 +21,30 @@ export var AIProviderName;
     AIProviderName["SAGEMAKER"] = "sagemaker";
     AIProviderName["AUTO"] = "auto";
 })(AIProviderName || (AIProviderName = {}));
+/**
+ * Popular Models for OpenRouter (300+ available at openrouter.ai/models)
+ * OpenRouter uses 'provider/model' format
+ */
+export var OpenRouterModels;
+(function (OpenRouterModels) {
+    // Anthropic Claude models
+    OpenRouterModels["CLAUDE_3_5_SONNET"] = "anthropic/claude-3-5-sonnet";
+    OpenRouterModels["CLAUDE_3_5_HAIKU"] = "anthropic/claude-3-5-haiku";
+    OpenRouterModels["CLAUDE_3_OPUS"] = "anthropic/claude-3-opus";
+    // OpenAI models
+    OpenRouterModels["GPT_4O"] = "openai/gpt-4o";
+    OpenRouterModels["GPT_4O_MINI"] = "openai/gpt-4o-mini";
+    OpenRouterModels["GPT_4_TURBO"] = "openai/gpt-4-turbo";
+    // Google models
+    OpenRouterModels["GEMINI_2_0_FLASH"] = "google/gemini-2.0-flash";
+    OpenRouterModels["GEMINI_1_5_PRO"] = "google/gemini-1.5-pro";
+    // Meta Llama models
+    OpenRouterModels["LLAMA_3_1_70B"] = "meta-llama/llama-3.1-70b-instruct";
+    OpenRouterModels["LLAMA_3_1_8B"] = "meta-llama/llama-3.1-8b-instruct";
+    // Mistral models
+    OpenRouterModels["MISTRAL_LARGE"] = "mistralai/mistral-large";
+    OpenRouterModels["MIXTRAL_8X7B"] = "mistralai/mixtral-8x7b-instruct";
+})(OpenRouterModels || (OpenRouterModels = {}));
 /**
  * Supported Models for Amazon Bedrock
  */

package/dist/core/baseProvider.d.ts

@@ -7,6 +7,7 @@ import type { MiddlewareFactoryOptions } from "../types/middlewareTypes.js";
 import type { StreamOptions, StreamResult } from "../types/streamTypes.js";
 import type { UnknownRecord } from "../types/common.js";
 import type { NeuroLink } from "../neurolink.js";
+import { TelemetryHandler } from "./modules/TelemetryHandler.js";
 /**
  * Abstract base class for all AI providers
  * Tools are integrated as first-class citizens - always available by default
@@ -26,7 +27,7 @@ export declare abstract class BaseProvider implements AIProvider {
     private readonly messageBuilder;
     private readonly streamHandler;
     private readonly generationHandler;
-    private readonly telemetryHandler;
+    protected readonly telemetryHandler: TelemetryHandler;
     private readonly utilities;
     private readonly toolsManager;
     constructor(modelName?: string, providerName?: AIProviderName, neurolink?: NeuroLink, middleware?: MiddlewareFactoryOptions);
@@ -260,13 +261,4 @@ export declare abstract class BaseProvider implements AIProvider {
      * @returns Array of prompt chunks
      */
     static chunkPrompt(prompt: string, maxChunkSize?: number, overlap?: number): string[];
-    /**
-     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
-     * This enables automatic OpenTelemetry tracing when telemetry is enabled
-     */
-    protected getStreamTelemetryConfig(options: StreamOptions | TextGenerationOptions, operationType?: "stream" | "generate"): {
-        isEnabled: boolean;
-        functionId?: string;
-        metadata?: Record<string, string | number | boolean>;
-    } | undefined;
 }

package/dist/core/baseProvider.js

@@ -48,7 +48,7 @@ export class BaseProvider {
         // Initialize composition modules
         this.messageBuilder = new MessageBuilder(this.providerName, this.modelName);
         this.streamHandler = new StreamHandler(this.providerName, this.modelName);
-        this.generationHandler = new GenerationHandler(this.providerName, this.modelName, () => this.supportsTools(), (options, type) => this.getStreamTelemetryConfig(options, type), (toolCalls, toolResults, options, timestamp) => this.handleToolExecutionStorage(toolCalls, toolResults, options, timestamp));
+        this.generationHandler = new GenerationHandler(this.providerName, this.modelName, () => this.supportsTools(), (options, type) => this.telemetryHandler.getTelemetryConfig(options, type), (toolCalls, toolResults, options, timestamp) => this.handleToolExecutionStorage(toolCalls, toolResults, options, timestamp));
         this.telemetryHandler = new TelemetryHandler(this.providerName, this.modelName, this.neurolink);
         this.utilities = new Utilities(this.providerName, this.modelName, this.defaultTimeout, this.middlewareOptions);
         this.toolsManager = new ToolsManager(this.providerName, this.directTools, this.neurolink, {
@@ -726,40 +726,4 @@ export class BaseProvider {
         }
         return chunks;
     }
-    /**
-     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
-     * This enables automatic OpenTelemetry tracing when telemetry is enabled
-     */
-    getStreamTelemetryConfig(options, operationType = "stream") {
-        // Check if telemetry is enabled via NeuroLink observability config
-        if (!this.neurolink?.isTelemetryEnabled()) {
-            return undefined;
-        }
-        const context = options.context;
-        const traceName = context?.traceName;
-        const userId = context?.userId;
-        const functionId = traceName ? traceName : userId ? userId : "guest";
-        const metadata = {
-            provider: this.providerName,
-            model: this.modelName,
-            toolsEnabled: !options.disableTools,
-            neurolink: true,
-            operationType,
-            originalProvider: this.providerName,
-        };
-        // Add sessionId if available
-        if ("sessionId" in options && options.sessionId) {
-            const sessionId = options.sessionId;
-            if (typeof sessionId === "string" ||
-                typeof sessionId === "number" ||
-                typeof sessionId === "boolean") {
-                metadata.sessionId = sessionId;
-            }
-        }
-        return {
-            isEnabled: true,
-            functionId,
-            metadata,
-        };
-    }
 }

package/dist/core/modules/TelemetryHandler.d.ts

@@ -50,9 +50,10 @@ export declare class TelemetryHandler {
         totalTokens?: number;
     }): Promise<number>;
     /**
-     * Get telemetry configuration for streaming/generation
+     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
+     * This enables automatic OpenTelemetry tracing when telemetry is enabled
      */
-    getStreamTelemetryConfig(options: StreamOptions | TextGenerationOptions, operationType?: "stream" | "generate"): {
+    getTelemetryConfig(options: StreamOptions | TextGenerationOptions, operationType?: "stream" | "generate"): {
         isEnabled: boolean;
         functionId?: string;
         metadata?: Record<string, string | number | boolean>;

package/dist/core/modules/TelemetryHandler.js

@@ -113,9 +113,10 @@ export class TelemetryHandler {
         }
     }
     /**
-     * Get telemetry configuration for streaming/generation
+     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
+     * This enables automatic OpenTelemetry tracing when telemetry is enabled
      */
-    getStreamTelemetryConfig(options, operationType = "stream") {
+    getTelemetryConfig(options, operationType = "stream") {
         // Check if telemetry is enabled via NeuroLink observability config
         if (!this.neurolink?.isTelemetryEnabled()) {
             return undefined;
@@ -125,6 +126,7 @@ export class TelemetryHandler {
         const userId = context?.userId;
         const functionId = traceName ? traceName : userId ? userId : "guest";
         const metadata = {
+            ...(context?.metadata || {}),
             provider: this.providerName,
             model: this.modelName,
             toolsEnabled: !options.disableTools,

package/dist/factories/providerRegistry.js

@@ -81,7 +81,12 @@ export class ProviderRegistry {
                 const { OpenAICompatibleProvider } = await import("../providers/openaiCompatible.js");
                 return new OpenAICompatibleProvider(modelName, sdk);
             }, process.env.OPENAI_COMPATIBLE_MODEL || undefined, // Enable auto-discovery when no model specified
-            ["openai-compatible", "openrouter", "vllm", "compatible"]);
+            ["openai-compatible", "vllm", "compatible"]);
+            // Register OpenRouter provider (300+ models from 60+ providers)
+            ProviderFactory.registerProvider(AIProviderName.OPENROUTER, async (modelName, _providerName, sdk) => {
+                const { OpenRouterProvider } = await import("../providers/openRouter.js");
+                return new OpenRouterProvider(modelName, sdk);
+            }, process.env.OPENROUTER_MODEL || "anthropic/claude-3-5-sonnet", ["openrouter", "or"]);
             // Register Amazon SageMaker provider
             ProviderFactory.registerProvider(AIProviderName.SAGEMAKER, async (modelName, _providerName, _sdk, region) => {
                 const { AmazonSageMakerProvider } = await import("../providers/amazonSagemaker.js");

package/dist/lib/adapters/providerImageAdapter.js

@@ -37,6 +37,7 @@ const IMAGE_LIMITS = {
     mistral: 10, // Conservative limit for Mistral
     // Note: Bedrock limit defined for future use when vision support is added
     bedrock: 20, // Same as Anthropic for Claude models on Bedrock
+    openrouter: 10, // Conservative limit, routes to various underlying providers
 };
 /**
  * Vision capability definitions for each provider
@@ -231,6 +232,37 @@ const VISION_CAPABILITIES = {
         // Groq models via LiteLLM (vision)
         "groq/llama-3.2-11b-vision-preview",
     ],
+    openrouter: [
+        // OpenRouter provides access to vision-capable models from multiple providers
+        // Anthropic Claude models (via OpenRouter)
+        "anthropic/claude-3-5-sonnet",
+        "anthropic/claude-3-5-haiku",
+        "anthropic/claude-3-opus",
+        "anthropic/claude-3-sonnet",
+        "anthropic/claude-3-haiku",
+        // OpenAI models (via OpenRouter)
+        "openai/gpt-4o",
+        "openai/gpt-4o-mini",
+        "openai/gpt-4-turbo",
+        "openai/gpt-4-vision-preview",
+        // Google models (via OpenRouter)
+        "google/gemini-2.5-pro",
+        "google/gemini-2.5-flash",
+        "google/gemini-2.0-flash",
+        "google/gemini-2.0-flash-001",
+        "google/gemini-1.5-pro",
+        "google/gemini-1.5-flash",
+        "google/gemini-pro-vision",
+        // Meta Llama models (vision-capable via OpenRouter)
+        "meta-llama/llama-3.2-90b-vision-instruct",
+        "meta-llama/llama-3.2-11b-vision-instruct",
+        // Pixtral/Mistral models (via OpenRouter)
+        "mistralai/pixtral-12b",
+        "mistralai/pixtral-large",
+        // Qwen models (via OpenRouter)
+        "qwen/qwen-2-vl-72b-instruct",
+        "qwen/qwen-2-vl-7b-instruct",
+    ],
     mistral: [
         // Mistral Large (latest has vision via Pixtral integration)
         "mistral-large-latest",
@@ -430,6 +462,11 @@ export class ProviderImageAdapter {
                     this.validateImageCount(images.length, "bedrock");
                     adaptedPayload = this.formatForAnthropic(text, images, true);
                     break;
+                case "openrouter":
+                    // OpenRouter routes to underlying providers, use OpenAI format
+                    this.validateImageCount(images.length, "openrouter");
+                    adaptedPayload = this.formatForOpenAI(text, images);
+                    break;
                 default:
                     throw new Error(`Vision not supported for provider: ${provider}`);
             }

package/dist/lib/constants/enums.d.ts

@@ -5,6 +5,7 @@ export declare enum AIProviderName {
     BEDROCK = "bedrock",
     OPENAI = "openai",
     OPENAI_COMPATIBLE = "openai-compatible",
+    OPENROUTER = "openrouter",
     VERTEX = "vertex",
     ANTHROPIC = "anthropic",
     AZURE = "azure",
@@ -16,6 +17,24 @@ export declare enum AIProviderName {
     SAGEMAKER = "sagemaker",
     AUTO = "auto"
 }
+/**
+ * Popular Models for OpenRouter (300+ available at openrouter.ai/models)
+ * OpenRouter uses 'provider/model' format
+ */
+export declare enum OpenRouterModels {
+    CLAUDE_3_5_SONNET = "anthropic/claude-3-5-sonnet",
+    CLAUDE_3_5_HAIKU = "anthropic/claude-3-5-haiku",
+    CLAUDE_3_OPUS = "anthropic/claude-3-opus",
+    GPT_4O = "openai/gpt-4o",
+    GPT_4O_MINI = "openai/gpt-4o-mini",
+    GPT_4_TURBO = "openai/gpt-4-turbo",
+    GEMINI_2_0_FLASH = "google/gemini-2.0-flash",
+    GEMINI_1_5_PRO = "google/gemini-1.5-pro",
+    LLAMA_3_1_70B = "meta-llama/llama-3.1-70b-instruct",
+    LLAMA_3_1_8B = "meta-llama/llama-3.1-8b-instruct",
+    MISTRAL_LARGE = "mistralai/mistral-large",
+    MIXTRAL_8X7B = "mistralai/mixtral-8x7b-instruct"
+}
 /**
  * Supported Models for Amazon Bedrock
  */

package/dist/lib/constants/enums.js

@@ -9,6 +9,7 @@ export var AIProviderName;
     AIProviderName["BEDROCK"] = "bedrock";
     AIProviderName["OPENAI"] = "openai";
     AIProviderName["OPENAI_COMPATIBLE"] = "openai-compatible";
+    AIProviderName["OPENROUTER"] = "openrouter";
     AIProviderName["VERTEX"] = "vertex";
     AIProviderName["ANTHROPIC"] = "anthropic";
     AIProviderName["AZURE"] = "azure";
@@ -20,6 +21,30 @@ export var AIProviderName;
     AIProviderName["SAGEMAKER"] = "sagemaker";
     AIProviderName["AUTO"] = "auto";
 })(AIProviderName || (AIProviderName = {}));
+/**
+ * Popular Models for OpenRouter (300+ available at openrouter.ai/models)
+ * OpenRouter uses 'provider/model' format
+ */
+export var OpenRouterModels;
+(function (OpenRouterModels) {
+    // Anthropic Claude models
+    OpenRouterModels["CLAUDE_3_5_SONNET"] = "anthropic/claude-3-5-sonnet";
+    OpenRouterModels["CLAUDE_3_5_HAIKU"] = "anthropic/claude-3-5-haiku";
+    OpenRouterModels["CLAUDE_3_OPUS"] = "anthropic/claude-3-opus";
+    // OpenAI models
+    OpenRouterModels["GPT_4O"] = "openai/gpt-4o";
+    OpenRouterModels["GPT_4O_MINI"] = "openai/gpt-4o-mini";
+    OpenRouterModels["GPT_4_TURBO"] = "openai/gpt-4-turbo";
+    // Google models
+    OpenRouterModels["GEMINI_2_0_FLASH"] = "google/gemini-2.0-flash";
+    OpenRouterModels["GEMINI_1_5_PRO"] = "google/gemini-1.5-pro";
+    // Meta Llama models
+    OpenRouterModels["LLAMA_3_1_70B"] = "meta-llama/llama-3.1-70b-instruct";
+    OpenRouterModels["LLAMA_3_1_8B"] = "meta-llama/llama-3.1-8b-instruct";
+    // Mistral models
+    OpenRouterModels["MISTRAL_LARGE"] = "mistralai/mistral-large";
+    OpenRouterModels["MIXTRAL_8X7B"] = "mistralai/mixtral-8x7b-instruct";
+})(OpenRouterModels || (OpenRouterModels = {}));
 /**
  * Supported Models for Amazon Bedrock
  */

package/dist/lib/core/baseProvider.d.ts

@@ -7,6 +7,7 @@ import type { MiddlewareFactoryOptions } from "../types/middlewareTypes.js";
 import type { StreamOptions, StreamResult } from "../types/streamTypes.js";
 import type { UnknownRecord } from "../types/common.js";
 import type { NeuroLink } from "../neurolink.js";
+import { TelemetryHandler } from "./modules/TelemetryHandler.js";
 /**
  * Abstract base class for all AI providers
  * Tools are integrated as first-class citizens - always available by default
@@ -26,7 +27,7 @@ export declare abstract class BaseProvider implements AIProvider {
     private readonly messageBuilder;
     private readonly streamHandler;
     private readonly generationHandler;
-    private readonly telemetryHandler;
+    protected readonly telemetryHandler: TelemetryHandler;
     private readonly utilities;
     private readonly toolsManager;
     constructor(modelName?: string, providerName?: AIProviderName, neurolink?: NeuroLink, middleware?: MiddlewareFactoryOptions);
@@ -260,13 +261,4 @@ export declare abstract class BaseProvider implements AIProvider {
      * @returns Array of prompt chunks
      */
     static chunkPrompt(prompt: string, maxChunkSize?: number, overlap?: number): string[];
-    /**
-     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
-     * This enables automatic OpenTelemetry tracing when telemetry is enabled
-     */
-    protected getStreamTelemetryConfig(options: StreamOptions | TextGenerationOptions, operationType?: "stream" | "generate"): {
-        isEnabled: boolean;
-        functionId?: string;
-        metadata?: Record<string, string | number | boolean>;
-    } | undefined;
 }

package/dist/lib/core/baseProvider.js

@@ -48,7 +48,7 @@ export class BaseProvider {
         // Initialize composition modules
         this.messageBuilder = new MessageBuilder(this.providerName, this.modelName);
         this.streamHandler = new StreamHandler(this.providerName, this.modelName);
-        this.generationHandler = new GenerationHandler(this.providerName, this.modelName, () => this.supportsTools(), (options, type) => this.getStreamTelemetryConfig(options, type), (toolCalls, toolResults, options, timestamp) => this.handleToolExecutionStorage(toolCalls, toolResults, options, timestamp));
+        this.generationHandler = new GenerationHandler(this.providerName, this.modelName, () => this.supportsTools(), (options, type) => this.telemetryHandler.getTelemetryConfig(options, type), (toolCalls, toolResults, options, timestamp) => this.handleToolExecutionStorage(toolCalls, toolResults, options, timestamp));
         this.telemetryHandler = new TelemetryHandler(this.providerName, this.modelName, this.neurolink);
         this.utilities = new Utilities(this.providerName, this.modelName, this.defaultTimeout, this.middlewareOptions);
         this.toolsManager = new ToolsManager(this.providerName, this.directTools, this.neurolink, {
@@ -726,41 +726,5 @@ export class BaseProvider {
         }
         return chunks;
     }
-    /**
-     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
-     * This enables automatic OpenTelemetry tracing when telemetry is enabled
-     */
-    getStreamTelemetryConfig(options, operationType = "stream") {
-        // Check if telemetry is enabled via NeuroLink observability config
-        if (!this.neurolink?.isTelemetryEnabled()) {
-            return undefined;
-        }
-        const context = options.context;
-        const traceName = context?.traceName;
-        const userId = context?.userId;
-        const functionId = traceName ? traceName : userId ? userId : "guest";
-        const metadata = {
-            provider: this.providerName,
-            model: this.modelName,
-            toolsEnabled: !options.disableTools,
-            neurolink: true,
-            operationType,
-            originalProvider: this.providerName,
-        };
-        // Add sessionId if available
-        if ("sessionId" in options && options.sessionId) {
-            const sessionId = options.sessionId;
-            if (typeof sessionId === "string" ||
-                typeof sessionId === "number" ||
-                typeof sessionId === "boolean") {
-                metadata.sessionId = sessionId;
-            }
-        }
-        return {
-            isEnabled: true,
-            functionId,
-            metadata,
-        };
-    }
 }
 //# sourceMappingURL=baseProvider.js.map

package/dist/lib/core/modules/TelemetryHandler.d.ts

@@ -50,9 +50,10 @@ export declare class TelemetryHandler {
         totalTokens?: number;
     }): Promise<number>;
     /**
-     * Get telemetry configuration for streaming/generation
+     * Create telemetry configuration for Vercel AI SDK experimental_telemetry
+     * This enables automatic OpenTelemetry tracing when telemetry is enabled
      */
-    getStreamTelemetryConfig(options: StreamOptions | TextGenerationOptions, operationType?: "stream" | "generate"): {
+    getTelemetryConfig(options: StreamOptions | TextGenerationOptions, operationType?: "stream" | "generate"): {
         isEnabled: boolean;
         functionId?: string;
         metadata?: Record<string, string | number | boolean>;