feat: add OpenAI provider with structured output support (#28)

* feat: add OpenAI provider with structured output support

Co-Authored-By: Han Xiao <han.xiao@jina.ai>

* fix: add @ai-sdk/openai dependency and fix modelConfigs access

Co-Authored-By: Han Xiao <han.xiao@jina.ai>

* fix: correct indentation in agent.ts

Co-Authored-By: Han Xiao <han.xiao@jina.ai>

* refactor: centralize model initialization in config.ts

Co-Authored-By: Han Xiao <han.xiao@jina.ai>

* refactor: improve model config access patterns

Co-Authored-By: Han Xiao <han.xiao@jina.ai>

* fix: remove unused imports

Co-Authored-By: Han Xiao <han.xiao@jina.ai>

* refactor: clean up

---------

Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: Han Xiao <han.xiao@jina.ai>

Dockerfile

@@ -15,6 +15,7 @@ COPY . .
 
 # Set environment variables
 ENV GEMINI_API_KEY=${GEMINI_API_KEY}
+ENV OPENAI_API_KEY=${OPENAI_API_KEY}
 ENV JINA_API_KEY=${JINA_API_KEY}
 ENV BRAVE_API_KEY=${BRAVE_API_KEY}
 

README.md

@@ -25,12 +25,7 @@ flowchart LR
 
 ## Install
 
-We use gemini for llm, [jina reader](https://jina.ai/reader) for searching and reading webpages. 
-
 ```bash
-export GEMINI_API_KEY=...  # for gemini api, ask han
-export JINA_API_KEY=jina_...  # free jina api key, get from https://jina.ai/reader
-
 git clone https://github.com/jina-ai/node-DeepResearch.git
 cd node-DeepResearch
 npm install
@@ -39,7 +34,14 @@ npm install
 
 ## Usage
 
+We use Gemini/OpenAI for reasoning, [Jina Reader](https://jina.ai/reader) for searching and reading webpages, you can get a free API key with 1M tokens from jina.ai. 
+
 ```bash
+export GEMINI_API_KEY=...  # for gemini
+# export OPENAI_API_KEY=... # for openai
+# export LLM_PROVIDER=openai # for openai
+export JINA_API_KEY=jina_...  # free jina api key, get from https://jina.ai/reader
+
 npm run dev $QUERY
 ```
 

docker-compose.yml

@@ -7,6 +7,7 @@ services:
       dockerfile: Dockerfile
     environment:
       - GEMINI_API_KEY=${GEMINI_API_KEY}
+      - OPENAI_API_KEY=${OPENAI_API_KEY}
       - JINA_API_KEY=${JINA_API_KEY}
       - BRAVE_API_KEY=${BRAVE_API_KEY}
     ports:

package-lock.json

@@ -1,15 +1,16 @@
 {
-  "name": "agentic-search",
+  "name": "node-deepresearch",
   "version": "1.0.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
-      "name": "agentic-search",
+      "name": "node-deepresearch",
       "version": "1.0.0",
-      "license": "ISC",
+      "license": "Apache-2.0",
       "dependencies": {
         "@ai-sdk/google": "^1.0.0",
+        "@ai-sdk/openai": "^1.1.9",
         "@types/cors": "^2.8.17",
         "@types/express": "^5.0.0",
         "@types/node-fetch": "^2.6.12",
@@ -51,6 +52,22 @@
         "zod": "^3.0.0"
       }
     },
+    "node_modules/@ai-sdk/openai": {
+      "version": "1.1.9",
+      "resolved": "https://registry.npmjs.org/@ai-sdk/openai/-/openai-1.1.9.tgz",
+      "integrity": "sha512-t/CpC4TLipdbgBJTMX/otzzqzCMBSPQwUOkYPGbT/jyuC86F+YO9o+LS0Ty2pGUE1kyT+B3WmJ318B16ZCg4hw==",
+      "license": "Apache-2.0",
+      "dependencies": {
+        "@ai-sdk/provider": "1.0.7",
+        "@ai-sdk/provider-utils": "2.1.6"
+      },
+      "engines": {
+        "node": ">=18"
+      },
+      "peerDependencies": {
+        "zod": "^3.0.0"
+      }
+    },
     "node_modules/@ai-sdk/provider": {
       "version": "1.0.7",
       "resolved": "https://registry.npmjs.org/@ai-sdk/provider/-/provider-1.0.7.tgz",

package.json

@@ -19,6 +19,7 @@
   "description": "",
   "dependencies": {
     "@ai-sdk/google": "^1.0.0",
+    "@ai-sdk/openai": "^1.1.9",
     "@types/cors": "^2.8.17",
     "@types/express": "^5.0.0",
     "@types/node-fetch": "^2.6.12",

src/agent.ts

@@ -1,6 +1,6 @@
-import {createGoogleGenerativeAI} from '@ai-sdk/google';
 import {z} from 'zod';
 import {generateObject} from 'ai';
+import {getModel, getMaxTokens, SEARCH_PROVIDER, STEP_SLEEP} from "./config";
 import {readUrl} from "./tools/read";
 import {handleGenerateObjectError} from './utils/error-handling';
 import fs from 'fs/promises';
@@ -10,7 +10,6 @@ import {rewriteQuery} from "./tools/query-rewriter";
 import {dedupQueries} from "./tools/dedup";
 import {evaluateAnswer} from "./tools/evaluator";
 import {analyzeSteps} from "./tools/error-analyzer";
-import {SEARCH_PROVIDER, STEP_SLEEP, modelConfigs} from "./config";
 import {TokenTracker} from "./utils/token-tracker";
 import {ActionTracker} from "./utils/action-tracker";
 import {StepAction, AnswerAction} from "./types";
@@ -325,7 +324,7 @@ export async function getResponse(question: string, tokenBudget: number = 1_000_
       false
     );
 
-    const model = createGoogleGenerativeAI({apiKey: process.env.GEMINI_API_KEY})(modelConfigs.agent.model);
+    const model = getModel('agent');
     let object;
     let totalTokens = 0;
     try {
@@ -333,7 +332,7 @@ export async function getResponse(question: string, tokenBudget: number = 1_000_
         model,
         schema: getSchema(allowReflect, allowRead, allowAnswer, allowSearch),
         prompt,
-        maxTokens: modelConfigs.agent.maxTokens
+        maxTokens: getMaxTokens('agent')
       });
       object = result.object;
       totalTokens = result.usage?.totalTokens || 0;
@@ -671,7 +670,7 @@ You decided to think out of the box or cut from a completely different angle.`);
       true
     );
 
-    const model = createGoogleGenerativeAI({apiKey: process.env.GEMINI_API_KEY})(modelConfigs.agentBeastMode.model);
+    const model = getModel('agentBeastMode');
     let object;
     let totalTokens = 0;
     try {
@@ -679,7 +678,7 @@ You decided to think out of the box or cut from a completely different angle.`);
         model,
         schema: getSchema(false, false, allowAnswer, false),
         prompt,
-        maxTokens: modelConfigs.agentBeastMode.maxTokens
+        maxTokens: getMaxTokens('agentBeastMode')
       });
       object = result.object;
       totalTokens = result.usage?.totalTokens || 0;

src/config.ts

@@ -1,13 +1,35 @@
 import dotenv from 'dotenv';
 import { ProxyAgent, setGlobalDispatcher } from 'undici';
+import { createGoogleGenerativeAI } from '@ai-sdk/google';
+import { createOpenAI } from '@ai-sdk/openai';
 
-interface ModelConfig {
+export type LLMProvider = 'openai' | 'gemini';
+export type ToolName = keyof ToolConfigs;
+
+function isValidProvider(provider: string): provider is LLMProvider {
+  return provider === 'openai' || provider === 'gemini';
+}
+
+function validateModelConfig(config: ModelConfig, toolName: string): ModelConfig {
+  if (typeof config.model !== 'string' || config.model.length === 0) {
+    throw new Error(`Invalid model name for ${toolName}`);
+  }
+  if (typeof config.temperature !== 'number' || config.temperature < 0 || config.temperature > 1) {
+    throw new Error(`Invalid temperature for ${toolName}`);
+  }
+  if (typeof config.maxTokens !== 'number' || config.maxTokens <= 0) {
+    throw new Error(`Invalid maxTokens for ${toolName}`);
+  }
+  return config;
+}
+
+export interface ModelConfig {
   model: string;
   temperature: number;
   maxTokens: number;
 }
 
-interface ToolConfigs {
+export interface ToolConfigs {
   dedup: ModelConfig;
   evaluator: ModelConfig;
   errorAnalyzer: ModelConfig;
@@ -31,44 +53,87 @@ if (process.env.https_proxy) {
 }
 
 export const GEMINI_API_KEY = process.env.GEMINI_API_KEY as string;
+export const OPENAI_API_KEY = process.env.OPENAI_API_KEY as string;
 export const JINA_API_KEY = process.env.JINA_API_KEY as string;
 export const BRAVE_API_KEY = process.env.BRAVE_API_KEY as string;
-export const SEARCH_PROVIDER: 'brave' | 'jina' | 'duck' = 'jina'
+export const SEARCH_PROVIDER: 'brave' | 'jina' | 'duck' = 'jina';
+export const LLM_PROVIDER: LLMProvider = (() => {
+  const provider = process.env.LLM_PROVIDER || 'gemini';
+  if (!isValidProvider(provider)) {
+    throw new Error(`Invalid LLM provider: ${provider}`);
+  }
+  return provider;
+})();
 
-const DEFAULT_MODEL = 'gemini-1.5-flash';
+const DEFAULT_GEMINI_MODEL = 'gemini-1.5-flash';
+const DEFAULT_OPENAI_MODEL = 'gpt-4o-mini';
 
-const defaultConfig: ModelConfig = {
-  model: DEFAULT_MODEL,
+const defaultGeminiConfig: ModelConfig = {
+  model: DEFAULT_GEMINI_MODEL,
   temperature: 0,
   maxTokens: 1000
 };
 
-export const modelConfigs: ToolConfigs = {
-  dedup: {
-    ...defaultConfig,
-    temperature: 0.1
+const defaultOpenAIConfig: ModelConfig = {
+  model: DEFAULT_OPENAI_MODEL,
+  temperature: 0,
+  maxTokens: 1000
+};
+
+export const modelConfigs: Record<LLMProvider, ToolConfigs> = {
+  gemini: {
+    dedup: validateModelConfig({ ...defaultGeminiConfig, temperature: 0.1 }, 'dedup'),
+    evaluator: validateModelConfig({ ...defaultGeminiConfig }, 'evaluator'),
+    errorAnalyzer: validateModelConfig({ ...defaultGeminiConfig }, 'errorAnalyzer'),
+    queryRewriter: validateModelConfig({ ...defaultGeminiConfig, temperature: 0.1 }, 'queryRewriter'),
+    agent: validateModelConfig({ ...defaultGeminiConfig, temperature: 0.7 }, 'agent'),
+    agentBeastMode: validateModelConfig({ ...defaultGeminiConfig, temperature: 0.7 }, 'agentBeastMode')
   },
-  evaluator: {
-    ...defaultConfig
-  },
-  errorAnalyzer: {
-    ...defaultConfig
-  },
-  queryRewriter: {
-    ...defaultConfig,
-    temperature: 0.1
-  },
-  agent: {
-    ...defaultConfig,
-    temperature: 0.7
-  },
-  agentBeastMode: {
-    ...defaultConfig,
-    temperature: 0.7
+  openai: {
+    dedup: validateModelConfig({ ...defaultOpenAIConfig, temperature: 0.1 }, 'dedup'),
+    evaluator: validateModelConfig({ ...defaultOpenAIConfig }, 'evaluator'),
+    errorAnalyzer: validateModelConfig({ ...defaultOpenAIConfig }, 'errorAnalyzer'),
+    queryRewriter: validateModelConfig({ ...defaultOpenAIConfig, temperature: 0.1 }, 'queryRewriter'),
+    agent: validateModelConfig({ ...defaultOpenAIConfig, temperature: 0.7 }, 'agent'),
+    agentBeastMode: validateModelConfig({ ...defaultOpenAIConfig, temperature: 0.7 }, 'agentBeastMode')
   }
 };
 
+export function getToolConfig(toolName: ToolName): ModelConfig {
+  if (!modelConfigs[LLM_PROVIDER][toolName]) {
+    throw new Error(`Invalid tool name: ${toolName}`);
+  }
+  return modelConfigs[LLM_PROVIDER][toolName];
+}
+
+export function getMaxTokens(toolName: ToolName): number {
+  return getToolConfig(toolName).maxTokens;
+}
+
+
+export function getModel(toolName: ToolName) {
+  const config = getToolConfig(toolName);
+
+  if (LLM_PROVIDER === 'openai') {
+    if (!OPENAI_API_KEY) {
+      throw new Error('OPENAI_API_KEY not found');
+    }
+    return createOpenAI({
+      apiKey: OPENAI_API_KEY,
+      compatibility: 'strict'
+    })(config.model);
+  }
+
+  if (!GEMINI_API_KEY) {
+    throw new Error('GEMINI_API_KEY not found');
+  }
+  return createGoogleGenerativeAI({ apiKey: GEMINI_API_KEY })(config.model);
+}
+
 export const STEP_SLEEP = 1000;
 
-if (!GEMINI_API_KEY) throw new Error("GEMINI_API_KEY not found");
+if (LLM_PROVIDER === 'gemini' && !GEMINI_API_KEY) throw new Error("GEMINI_API_KEY not found");
+if (LLM_PROVIDER === 'openai' && !OPENAI_API_KEY) throw new Error("OPENAI_API_KEY not found");
 if (!JINA_API_KEY) throw new Error("JINA_API_KEY not found");
+
+console.log('LLM Provider:', LLM_PROVIDER)

src/tools/__tests__/dedup.test.ts

@@ -1,8 +1,27 @@
 import { dedupQueries } from '../dedup';
+import { LLMProvider } from '../../config';
 
 describe('dedupQueries', () => {
+  const providers: Array<LLMProvider> = ['openai', 'gemini'];
+  const originalEnv = process.env;
+
+  beforeEach(() => {
+    jest.resetModules();
+    process.env = { ...originalEnv };
+  });
+
+  afterEach(() => {
+    process.env = originalEnv;
+  });
+
+  providers.forEach(provider => {
+    describe(`with ${provider} provider`, () => {
+      beforeEach(() => {
+        process.env.LLM_PROVIDER = provider;
+      });
+
       it('should remove duplicate queries', async () => {
-    jest.setTimeout(10000); // Increase timeout to 10s
+        jest.setTimeout(10000);
         const queries = ['typescript tutorial', 'typescript tutorial', 'javascript basics'];
         const { unique_queries } = await dedupQueries(queries, []);
         expect(unique_queries).toHaveLength(2);
@@ -14,3 +33,5 @@ describe('dedupQueries', () => {
         expect(unique_queries).toHaveLength(0);
       });
     });
+  });
+});

src/tools/__tests__/error-analyzer.test.ts

@@ -1,6 +1,25 @@
 import { analyzeSteps } from '../error-analyzer';
+import { LLMProvider } from '../../config';
 
 describe('analyzeSteps', () => {
+  const providers: Array<LLMProvider> = ['openai', 'gemini'];
+  const originalEnv = process.env;
+
+  beforeEach(() => {
+    jest.resetModules();
+    process.env = { ...originalEnv };
+  });
+
+  afterEach(() => {
+    process.env = originalEnv;
+  });
+
+  providers.forEach(provider => {
+    describe(`with ${provider} provider`, () => {
+      beforeEach(() => {
+        process.env.LLM_PROVIDER = provider;
+      });
+
       it('should analyze error steps', async () => {
         const { response } = await analyzeSteps(['Step 1: Search failed', 'Step 2: Invalid query']);
         expect(response).toHaveProperty('recap');
@@ -8,3 +27,5 @@ describe('analyzeSteps', () => {
         expect(response).toHaveProperty('improvement');
       });
     });
+  });
+});

src/tools/__tests__/evaluator.test.ts

@@ -1,7 +1,26 @@
 import { evaluateAnswer } from '../evaluator';
 import { TokenTracker } from '../../utils/token-tracker';
+import { LLMProvider } from '../../config';
 
 describe('evaluateAnswer', () => {
+  const providers: Array<LLMProvider> = ['openai', 'gemini'];
+  const originalEnv = process.env;
+
+  beforeEach(() => {
+    jest.resetModules();
+    process.env = { ...originalEnv };
+  });
+
+  afterEach(() => {
+    process.env = originalEnv;
+  });
+
+  providers.forEach(provider => {
+    describe(`with ${provider} provider`, () => {
+      beforeEach(() => {
+        process.env.LLM_PROVIDER = provider;
+      });
+
       it('should evaluate answer definitiveness', async () => {
         const tokenTracker = new TokenTracker();
         const { response } = await evaluateAnswer(
@@ -25,3 +44,5 @@ describe('evaluateAnswer', () => {
         expect(tokens).toBeGreaterThan(0);
       });
     });
+  });
+});

src/tools/__tests__/query-rewriter.test.ts

@@ -1,6 +1,25 @@
 import { rewriteQuery } from '../query-rewriter';
+import { LLMProvider } from '../../config';
 
 describe('rewriteQuery', () => {
+  const providers: Array<LLMProvider> = ['openai', 'gemini'];
+  const originalEnv = process.env;
+
+  beforeEach(() => {
+    jest.resetModules();
+    process.env = { ...originalEnv };
+  });
+
+  afterEach(() => {
+    process.env = originalEnv;
+  });
+
+  providers.forEach(provider => {
+    describe(`with ${provider} provider`, () => {
+      beforeEach(() => {
+        process.env.LLM_PROVIDER = provider;
+      });
+
       it('should rewrite search query', async () => {
         const { queries } = await rewriteQuery({
           action: 'search',
@@ -11,3 +30,5 @@ describe('rewriteQuery', () => {
         expect(queries.length).toBeGreaterThan(0);
       });
     });
+  });
+});

src/tools/dedup.ts

@@ -1,11 +1,11 @@
-import { createGoogleGenerativeAI } from '@ai-sdk/google';
 import { z } from 'zod';
 import { generateObject } from 'ai';
-import { modelConfigs } from "../config";
+import { getModel, getMaxTokens } from "../config";
 import { TokenTracker } from "../utils/token-tracker";
 import { handleGenerateObjectError } from '../utils/error-handling';
 import type { DedupResponse } from '../types';
 
+const model = getModel('dedup');
 
 const responseSchema = z.object({
   think: z.string().describe('Strategic reasoning about the overall deduplication approach'),
@@ -13,8 +13,6 @@ const responseSchema = z.object({
     .describe('Array of semantically unique queries').max(3)
 });
 
-const model = createGoogleGenerativeAI({ apiKey: process.env.GEMINI_API_KEY })(modelConfigs.dedup.model);
-
 function getPrompt(newQueries: string[], existingQueries: string[]): string {
   return `You are an expert in semantic similarity analysis. Given a set of queries (setA) and a set of queries (setB)
 
@@ -77,7 +75,7 @@ export async function dedupQueries(newQueries: string[], existingQueries: string
         model,
         schema: responseSchema,
         prompt,
-        maxTokens: modelConfigs.dedup.maxTokens
+        maxTokens: getMaxTokens('dedup')
       });
       object = result.object;
       tokens = result.usage?.totalTokens || 0;

src/tools/error-analyzer.ts

@@ -1,18 +1,19 @@
-import { createGoogleGenerativeAI } from '@ai-sdk/google';
 import { z } from 'zod';
 import { generateObject } from 'ai';
-import { modelConfigs } from "../config";
+import { getModel, getMaxTokens } from "../config";
 import { TokenTracker } from "../utils/token-tracker";
 import { ErrorAnalysisResponse } from '../types';
 import { handleGenerateObjectError } from '../utils/error-handling';
 
+const model = getModel('errorAnalyzer');
+
 const responseSchema = z.object({
   recap: z.string().describe('Recap of the actions taken and the steps conducted'),
   blame: z.string().describe('Which action or the step was the root cause of the answer rejection'),
   improvement: z.string().describe('Suggested key improvement for the next iteration, do not use bullet points, be concise and hot-take vibe.')
 });
 
-const model = createGoogleGenerativeAI({ apiKey: process.env.GEMINI_API_KEY })(modelConfigs.errorAnalyzer.model);
+
 
 function getPrompt(diaryContext: string[]): string {
   return `You are an expert at analyzing search and reasoning processes. Your task is to analyze the given sequence of steps and identify what went wrong in the search process.
@@ -112,7 +113,7 @@ export async function analyzeSteps(diaryContext: string[], tracker?: TokenTracke
         model,
         schema: responseSchema,
         prompt,
-        maxTokens: modelConfigs.errorAnalyzer.maxTokens
+        maxTokens: getMaxTokens('errorAnalyzer')
       });
       object = result.object;
       tokens = result.usage?.totalTokens || 0;

src/tools/evaluator.ts

@@ -1,17 +1,18 @@
-import { createGoogleGenerativeAI } from '@ai-sdk/google';
 import { z } from 'zod';
 import { generateObject } from 'ai';
-import { modelConfigs } from "../config";
+import { getModel, getMaxTokens } from "../config";
 import { TokenTracker } from "../utils/token-tracker";
 import { EvaluationResponse } from '../types';
 import { handleGenerateObjectError } from '../utils/error-handling';
 
+const model = getModel('evaluator');
+
 const responseSchema = z.object({
   is_definitive: z.boolean().describe('Whether the answer provides a definitive response without uncertainty or \'I don\'t know\' type statements'),
   reasoning: z.string().describe('Explanation of why the answer is or isn\'t definitive')
 });
 
-const model = createGoogleGenerativeAI({ apiKey: process.env.GEMINI_API_KEY })(modelConfigs.evaluator.model);
+
 
 function getPrompt(question: string, answer: string): string {
   return `You are an evaluator of answer definitiveness. Analyze if the given answer provides a definitive response or not.
@@ -57,7 +58,7 @@ export async function evaluateAnswer(question: string, answer: string, tracker?:
         model,
         schema: responseSchema,
         prompt,
-        maxTokens: modelConfigs.evaluator.maxTokens
+        maxTokens: getMaxTokens('evaluator')
       });
       object = result.object;
       totalTokens = result.usage?.totalTokens || 0;

src/tools/query-rewriter.ts

@@ -1,11 +1,12 @@
-import { createGoogleGenerativeAI } from '@ai-sdk/google';
 import { z } from 'zod';
-import { modelConfigs } from "../config";
+import { generateObject } from 'ai';
+import { getModel, getMaxTokens } from "../config";
 import { TokenTracker } from "../utils/token-tracker";
 import { SearchAction, KeywordsResponse } from '../types';
-import { generateObject } from 'ai';
 import { handleGenerateObjectError } from '../utils/error-handling';
 
+const model = getModel('queryRewriter');
+
 const responseSchema = z.object({
   think: z.string().describe('Strategic reasoning about query complexity and search approach'),
   queries: z.array(z.string().describe('Search query, must be less than 30 characters'))
@@ -14,7 +15,7 @@ const responseSchema = z.object({
     .describe('Array of search queries, orthogonal to each other')
 });
 
-const model = createGoogleGenerativeAI({ apiKey: process.env.GEMINI_API_KEY })(modelConfigs.queryRewriter.model);
+
 
 function getPrompt(action: SearchAction): string {
   return `You are an expert Information Retrieval Assistant. Transform user queries into precise keyword combinations with strategic reasoning and appropriate search operators.
@@ -102,7 +103,7 @@ export async function rewriteQuery(action: SearchAction, tracker?: TokenTracker)
         model,
         schema: responseSchema,
         prompt,
-        maxTokens: modelConfigs.queryRewriter.maxTokens
+        maxTokens: getMaxTokens('queryRewriter')
       });
       object = result.object;
       tokens = result.usage?.totalTokens || 0;