diff --git a/.env.example b/.env.example
index b44266f..1dcecf2 100644
--- a/.env.example
+++ b/.env.example
@@ -31,13 +31,23 @@ REFRESH_INTERVAL_MINUTES=15
 
 # === LLM Layer (optional) ===
 # Enables AI-enhanced trade ideas and breaking news Telegram alerts.
-# Provider options: anthropic | openai | gemini | codex | openrouter | minimax | mistral
+# Provider options: anthropic | openai | gemini | codex | openrouter | minimax | mistral | ollama
 LLM_PROVIDER=
-# Not needed for codex (uses ~/.codex/auth.json)
+# Not needed for codex (uses ~/.codex/auth.json) or ollama (local, no key required)
 LLM_API_KEY=
 # Optional override. Each provider has a sensible default:
-# anthropic: claude-sonnet-4-6 | openai: gpt-5.4 | gemini: gemini-3.1-pro | codex: gpt-5.3-codex | openrouter: openrouter/auto | minimax: MiniMax-M2.5
+# anthropic: claude-sonnet-4-6 | openai: gpt-5.4 | gemini: gemini-3.1-pro | codex: gpt-5.3-codex | openrouter: openrouter/auto | minimax: MiniMax-M2.5 | ollama: llama3.2
 LLM_MODEL=
+# Ollama only — base URL of your Ollama instance (default: http://localhost:11434)
+OLLAMA_BASE_URL=
+
+# === Localization (optional) ===
+# Set UI language. Supported: en (English), fr (French), vi (Vietnamese)
+# Trade ideas and LLM output will also be generated in the selected language.
+CRUCIX_LANG=
+# Translate dynamic content (news/OSINT headlines) via LLM when lang != en
+# Requires LLM_PROVIDER to be configured. Default: true when lang is set.
+CRUCIX_TRANSLATE=true
 
 # === Telegram Alerts (optional, requires LLM) ===
 # Create a bot via @BotFather, get chat ID via @userinfobot
diff --git a/lib/llm/index.mjs b/lib/llm/index.mjs
index b2d16ee..7d54b3a 100644
--- a/lib/llm/index.mjs
+++ b/lib/llm/index.mjs
@@ -7,6 +7,7 @@ import { GeminiProvider } from './gemini.mjs';
 import { CodexProvider } from './codex.mjs';
 import { MiniMaxProvider } from './minimax.mjs';
 import { MistralProvider } from './mistral.mjs';
+import { OllamaProvider } from './ollama.mjs';
 
 export { LLMProvider } from './provider.mjs';
 export { AnthropicProvider } from './anthropic.mjs';
@@ -16,6 +17,7 @@ export { GeminiProvider } from './gemini.mjs';
 export { CodexProvider } from './codex.mjs';
 export { MiniMaxProvider } from './minimax.mjs';
 export { MistralProvider } from './mistral.mjs';
+export { OllamaProvider } from './ollama.mjs';
 
 /**
  * Create an LLM provider based on config.
@@ -42,6 +44,8 @@ export function createLLMProvider(llmConfig) {
       return new MiniMaxProvider({ apiKey, model });
     case 'mistral':
       return new MistralProvider({ apiKey, model });
+    case 'ollama':
+      return new OllamaProvider({ model, baseUrl: llmConfig.ollamaBaseUrl });
     default:
       console.warn(`[LLM] Unknown provider "${provider}". LLM features disabled.`);
       return null;
diff --git a/lib/llm/ollama.mjs b/lib/llm/ollama.mjs
new file mode 100644
index 0000000..76926a8
--- /dev/null
+++ b/lib/llm/ollama.mjs
@@ -0,0 +1,53 @@
+// Ollama Provider — local LLM via Ollama REST API (no API key required)
+// Docs: https://github.com/ollama/ollama/blob/main/docs/api.md
+
+import { LLMProvider } from './provider.mjs';
+
+export class OllamaProvider extends LLMProvider {
+  constructor(config) {
+    super(config);
+    this.name = 'ollama';
+    this.model = config.model || 'llama3.2';
+    this.baseUrl = (config.baseUrl || process.env.OLLAMA_BASE_URL || 'http://localhost:11434').replace(/\/$/, '');
+  }
+
+  get isConfigured() { return !!this.model; }
+
+  async complete(systemPrompt, userMessage, opts = {}) {
+    const url = `${this.baseUrl}/api/chat`;
+
+    const res = await fetch(url, {
+      method: 'POST',
+      headers: { 'Content-Type': 'application/json' },
+      body: JSON.stringify({
+        model: this.model,
+        stream: false,
+        options: {
+          num_predict: opts.maxTokens || 4096,
+        },
+        messages: [
+          { role: 'system', content: systemPrompt },
+          { role: 'user', content: userMessage },
+        ],
+      }),
+      signal: AbortSignal.timeout(opts.timeout || 120000), // Ollama local can be slower
+    });
+
+    if (!res.ok) {
+      const err = await res.text().catch(() => '');
+      throw new Error(`Ollama API ${res.status}: ${err.substring(0, 200)}`);
+    }
+
+    const data = await res.json();
+    const text = data.message?.content || '';
+
+    return {
+      text,
+      usage: {
+        inputTokens: data.prompt_eval_count || 0,
+        outputTokens: data.eval_count || 0,
+      },
+      model: data.model || this.model,
+    };
+  }
+}