feat: add Requesty as an OpenAI-compatible LLM provider

README.md

@@ -614,6 +614,7 @@ curl -X POST http://localhost:21100/api/v1/ingest \
 | **DashScope** | qwen-plus, qwen-turbo | 通义千问, OpenAI-compatible |
 | **Ollama** | qwen2.5, llama3.2 | Fully local, zero cost |
 | **OpenRouter** | Any of 100+ models | Unified gateway |
+| **Requesty** | Any of 300+ models | OpenAI-compatible gateway |
 
 Each extraction/lifecycle LLM can be configured with a primary provider, an optional fallback provider, retry settings, and per-provider timeout.
 

packages/dashboard/src/pages/Settings/types.ts

@@ -67,6 +67,21 @@ export const LLM_PROVIDERS: Record<string, ProviderPreset> = {
     ],
     envKey: 'OPENROUTER_API_KEY',
   },
+  requesty: {
+    label: 'Requesty',
+    defaultBaseUrl: 'https://router.requesty.ai/v1',
+    models: [
+      'openai/gpt-4o-mini',
+      'openai/gpt-5.2',
+      'anthropic/claude-haiku-4-5',
+      'anthropic/claude-sonnet-4-6',
+      'google/gemini-2.5-flash',
+      'google/gemini-2.5-pro',
+      'deepseek/deepseek-chat',
+      'meta-llama/llama-4-maverick',
+    ],
+    envKey: 'REQUESTY_API_KEY',
+  },
   ollama: {
     label: 'Ollama (Local)',
     defaultBaseUrl: 'http://localhost:11434',

packages/server/src/llm/cascade.ts

@@ -4,6 +4,7 @@ import { AnthropicLLMProvider } from './anthropic.js';
 import { OllamaLLMProvider } from './ollama.js';
 import { GoogleLLMProvider } from './google.js';
 import { OpenRouterLLMProvider } from './openrouter.js';
+import { RequestyLLMProvider } from './requesty.js';
 import { DeepSeekLLMProvider } from './deepseek.js';
 import { normalizeLLMRetryConfig } from './config-utils.js';
 import { createLogger } from '../utils/logger.js';
@@ -144,6 +145,8 @@ export function createLLMProvider(config: LLMProviderConfig): LLMProvider {
       });
     case 'openrouter':
       return new OpenRouterLLMProvider(config);
+    case 'requesty':
+      return new RequestyLLMProvider(config);
     case 'ollama':
       return new OllamaLLMProvider(config);
     case 'none':

packages/server/src/llm/requesty.ts

@@ -0,0 +1,58 @@
+import type { LLMProvider, LLMCompletionOpts } from './interface.js';
+import { createLogger } from '../utils/logger.js';
+
+const log = createLogger('llm-requesty');
+
+/**
+ * Requesty LLM Provider — routes to any model via Requesty's unified API.
+ * Uses OpenAI-compatible format.
+ */
+export class RequestyLLMProvider implements LLMProvider {
+  readonly name = 'requesty';
+  private apiKey: string;
+  private model: string;
+  private baseUrl: string;
+  private timeoutMs: number;
+
+  constructor(opts: { apiKey?: string; model?: string; baseUrl?: string; timeoutMs?: number }) {
+    this.apiKey = opts.apiKey || process.env.REQUESTY_API_KEY || '';
+    this.model = opts.model || 'openai/gpt-4o-mini';
+    this.baseUrl = (opts.baseUrl || 'https://router.requesty.ai/v1').replace(/\/+$/, '');
+    this.timeoutMs = opts.timeoutMs ?? 30000;
+  }
+
+  async complete(prompt: string, opts?: LLMCompletionOpts): Promise<string> {
+    if (!this.apiKey) throw new Error('Requesty API key not configured');
+
+    const messages: any[] = [];
+    if (opts?.systemPrompt) {
+      messages.push({ role: 'system', content: opts.systemPrompt });
+    }
+    messages.push({ role: 'user', content: prompt });
+
+    const res = await fetch(`${this.baseUrl}/chat/completions`, {
+      method: 'POST',
+      headers: {
+        'Content-Type': 'application/json',
+        Authorization: `Bearer ${this.apiKey}`,
+        'HTTP-Referer': 'https://github.com/rikouu/cortex',
+        'X-Title': 'Cortex Memory Service',
+      },
+      body: JSON.stringify({
+        model: this.model,
+        messages,
+        max_tokens: opts?.maxTokens || 500,
+        temperature: opts?.temperature ?? 0.3,
+      }),
+      signal: AbortSignal.timeout(this.timeoutMs),
+    });
+
+    if (!res.ok) {
+      const body = await res.text();
+      throw new Error(`Requesty API error ${res.status}: ${body}`);
+    }
+
+    const data = (await res.json()) as any;
+    return data.choices?.[0]?.message?.content || '';
+  }
+}

packages/server/src/utils/config.ts

@@ -9,7 +9,7 @@ const LLMRetrySchema = z.object({
 });
 
 const LLMProviderSchema = z.object({
-  provider: z.enum(['openai', 'anthropic', 'google', 'gemini', 'deepseek', 'dashscope', 'openrouter', 'ollama', 'none']),
+  provider: z.enum(['openai', 'anthropic', 'google', 'gemini', 'deepseek', 'dashscope', 'openrouter', 'requesty', 'ollama', 'none']),
   model: z.string().optional(),
   apiKey: z.string().optional(),
   baseUrl: z.string().optional(),