codex-linux 1.0.0

package/src/main/providers/FreeModelsProvider.ts

@@ -0,0 +1,1156 @@
+import OpenAI from 'openai';
+import log from 'electron-log';
+import { ProviderConfig, ProviderModel } from '../../shared/types';
+
+interface FreeModelConfig {
+  provider: string;
+  baseUrl: string;
+  requiresApiKey: boolean;
+  defaultApiKey?: string;
+  description: string;
+}
+
+const FREE_MODEL_CONFIGS: Record<string, FreeModelConfig> = {
+  openrouter: {
+    provider: 'OpenRouter',
+    baseUrl: 'https://openrouter.ai/api/v1',
+    requiresApiKey: false,
+    defaultApiKey: '',
+    description: 'Free models via OpenRouter (no API key required for free tier)'
+  },
+  ollama: {
+    provider: 'Ollama',
+    baseUrl: 'http://localhost:11434/v1',
+    requiresApiKey: false,
+    defaultApiKey: '',
+    description: 'Local Ollama server - runs models locally on your machine'
+  },
+  nvidia: {
+    provider: 'NVIDIA NIM',
+    baseUrl: 'https://integrate.api.nvidia.com/v1',
+    requiresApiKey: true,
+    description: 'NVIDIA NIM API - Free tier with phone verification (40 req/min)'
+  },
+  groq: {
+    provider: 'Groq',
+    baseUrl: 'https://api.groq.com/openai/v1',
+    requiresApiKey: true,
+    description: 'Fast inference with Llama/Mixtral models (free tier available)'
+  },
+  google: {
+    provider: 'Google AI Studio',
+    baseUrl: 'https://generativelanguage.googleapis.com/v1beta/openai',
+    requiresApiKey: true,
+    description: 'Gemini models free tier (requires Google AI Studio API key)'
+  },
+  cerebras: {
+    provider: 'Cerebras',
+    baseUrl: 'https://api.cerebras.ai/v1',
+    requiresApiKey: true,
+    description: 'Ultra-fast inference (free tier available)'
+  },
+  cloudflare: {
+    provider: 'Cloudflare Workers AI',
+    baseUrl: 'https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/v1',
+    requiresApiKey: true,
+    description: 'Edge AI with free tier (10,000 neurons/day)'
+  }
+};
+
+const FREE_MODELS: ProviderModel[] = [
+  {
+    id: 'meta-llama/llama-3.3-70b-instruct:free',
+    name: 'Llama 3.3 70B (Free)',
+    description: 'Meta Llama 3.3 70B - GPT-4 level performance, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'google/gemini-2.0-flash-exp:free',
+    name: 'Gemini 2.0 Flash (Free)',
+    description: 'Google Gemini 2.0 Flash - 1M context, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 1000000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'google/gemini-2.5-flash-preview:free',
+    name: 'Gemini 2.5 Flash Preview (Free)',
+    description: 'Google Gemini 2.5 Flash Preview - via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 1000000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'google/gemma-3-27b-it:free',
+    name: 'Gemma 3 27B (Free)',
+    description: 'Google Gemma 3 27B Instruct, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'google/gemma-3-12b-it:free',
+    name: 'Gemma 3 12B (Free)',
+    description: 'Google Gemma 3 12B Instruct, via OpenRouter',
+    maxTokens: 4096,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'mistralai/mistral-small-3.1-24b-instruct:free',
+    name: 'Mistral Small 3.1 24B (Free)',
+    description: 'Mistral Small 3.1 24B Instruct, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'deepseek/deepseek-r1-0528:free',
+    name: 'DeepSeek R1 (Free)',
+    description: 'DeepSeek R1 reasoning model, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'qwen/qwen3-4b:free',
+    name: 'Qwen 3 4B (Free)',
+    description: 'Alibaba Qwen 3 4B, via OpenRouter',
+    maxTokens: 4096,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'qwen/qwen3-coder:free',
+    name: 'Qwen 3 Coder (Free)',
+    description: 'Alibaba Qwen 3 Coder - optimized for code, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'qwen/qwen3-next-80b-a3b-instruct:free',
+    name: 'Qwen 3 Next 80B (Free)',
+    description: 'Alibaba Qwen 3 Next 80B, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'nvidia/nemotron-nano-12b-v2-vl:free',
+    name: 'Nemotron Nano 12B VL (Free)',
+    description: 'NVIDIA Nemotron Nano 12B Vision Language, via OpenRouter',
+    maxTokens: 4096,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'nousresearch/hermes-3-llama-3.1-405b:free',
+    name: 'Hermes 3 Llama 405B (Free)',
+    description: 'Nous Hermes 3 Llama 3.1 405B, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'openai/gpt-oss-120b:free',
+    name: 'GPT-OSS 120B (Free)',
+    description: 'OpenAI OSS 120B model, via OpenRouter',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'z-ai/glm-4.5-air:free',
+    name: 'GLM 4.5 Air (Free)',
+    description: 'Z.ai GLM 4.5 Air, via OpenRouter',
+    maxTokens: 4096,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'liquid/lfm-2.5-1.2b-instruct:free',
+    name: 'LFM 2.5 1.2B (Free)',
+    description: 'Liquid LFM 2.5 1.2B Instruct, via OpenRouter',
+    maxTokens: 2048,
+    contextWindow: 8192,
+    supportsTools: false,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'upstage/solar-pro-3:free',
+    name: 'Solar Pro 3 (Free)',
+    description: 'Upstage Solar Pro 3, via OpenRouter',
+    maxTokens: 4096,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'openrouter'
+  },
+  {
+    id: 'moonshotai/kimi-k2.5',
+    name: 'Kimi K2.5 (NVIDIA)',
+    description: 'Moonshot Kimi K2.5 - Advanced reasoning, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'deepseek-ai/deepseek-r1',
+    name: 'DeepSeek R1 (NVIDIA)',
+    description: 'DeepSeek R1 reasoning model, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'meta/llama-3.3-70b-instruct',
+    name: 'Llama 3.3 70B (NVIDIA)',
+    description: 'Meta Llama 3.3 70B Instruct, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'meta/llama-3.1-405b-instruct',
+    name: 'Llama 3.1 405B (NVIDIA)',
+    description: 'Meta Llama 3.1 405B Instruct, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'meta/llama-3.1-8b-instruct',
+    name: 'Llama 3.1 8B (NVIDIA)',
+    description: 'Meta Llama 3.1 8B Instruct, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'mistralai/mistral-large',
+    name: 'Mistral Large (NVIDIA)',
+    description: 'Mistral Large model, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'mistralai/codestral-22b',
+    name: 'Codestral 22B (NVIDIA)',
+    description: 'Mistral Codestral - Code generation, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'mistralai/mathstral-7b-v0.1',
+    name: 'Mathstral 7B (NVIDIA)',
+    description: 'Mistral Mathstral - Math reasoning, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'nvidia/llama-3.1-nemotron-70b-ultra',
+    name: 'Nemotron 70B Ultra (NVIDIA)',
+    description: 'NVIDIA Nemotron 70B Ultra, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'nvidia/llama-3.1-nemotron-51b-instruct',
+    name: 'Nemotron 51B (NVIDIA)',
+    description: 'NVIDIA Nemotron 51B Instruct, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'google/gemma-2-27b-it',
+    name: 'Gemma 2 27B (NVIDIA)',
+    description: 'Google Gemma 2 27B Instruct, via NVIDIA NIM',
+    maxTokens: 8192,
+    contextWindow: 8192,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'google/gemma-2-9b-it',
+    name: 'Gemma 2 9B (NVIDIA)',
+    description: 'Google Gemma 2 9B Instruct, via NVIDIA NIM',
+    maxTokens: 8192,
+    contextWindow: 8192,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'google/gemma-2-2b-it',
+    name: 'Gemma 2 2B (NVIDIA)',
+    description: 'Google Gemma 2 2B Instruct, via NVIDIA NIM',
+    maxTokens: 8192,
+    contextWindow: 8192,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'qwen/qwen2.5-72b-instruct',
+    name: 'Qwen 2.5 72B (NVIDIA)',
+    description: 'Alibaba Qwen 2.5 72B Instruct, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'qwen/qwen2.5-math-72b-instruct',
+    name: 'Qwen 2.5 Math 72B (NVIDIA)',
+    description: 'Qwen 2.5 Math 72B - Math reasoning, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 4096,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'qwen/qwen2.5-coder-32b-instruct',
+    name: 'Qwen 2.5 Coder 32B (NVIDIA)',
+    description: 'Qwen 2.5 Coder - Code generation, via NVIDIA NIM',
+    maxTokens: 16384,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'microsoft/phi-3-medium-128k-instruct',
+    name: 'Phi-3 Medium 128K (NVIDIA)',
+    description: 'Microsoft Phi-3 Medium 128K context, via NVIDIA NIM',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'microsoft/phi-3-mini-128k-instruct',
+    name: 'Phi-3 Mini 128K (NVIDIA)',
+    description: 'Microsoft Phi-3 Mini 128K context, via NVIDIA NIM',
+    maxTokens: 4096,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'snowflake/arctic',
+    name: 'Arctic (NVIDIA)',
+    description: 'Snowflake Arctic - Enterprise LLM, via NVIDIA NIM',
+    maxTokens: 4096,
+    contextWindow: 4096,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'upstage/solar-10.7b-instruct',
+    name: 'Solar 10.7B (NVIDIA)',
+    description: 'Upstage Solar 10.7B Instruct, via NVIDIA NIM',
+    maxTokens: 4096,
+    contextWindow: 4096,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'nvidia'
+  },
+  {
+    id: 'llama3.2:latest',
+    name: 'Llama 3.2 (Ollama)',
+    description: 'Meta Llama 3.2 - Local via Ollama',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'llama3.2:1b',
+    name: 'Llama 3.2 1B (Ollama)',
+    description: 'Meta Llama 3.2 1B - Lightweight, via Ollama',
+    maxTokens: 4096,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'llama3.2:3b',
+    name: 'Llama 3.2 3B (Ollama)',
+    description: 'Meta Llama 3.2 3B - Compact, via Ollama',
+    maxTokens: 4096,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'llama3.3:latest',
+    name: 'Llama 3.3 70B (Ollama)',
+    description: 'Meta Llama 3.3 70B - Full size, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'llama3.1:8b',
+    name: 'Llama 3.1 8B (Ollama)',
+    description: 'Meta Llama 3.1 8B - Balanced, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'llama3.1:70b',
+    name: 'Llama 3.1 70B (Ollama)',
+    description: 'Meta Llama 3.1 70B - Large, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'mistral:latest',
+    name: 'Mistral (Ollama)',
+    description: 'Mistral 7B - Fast and efficient, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'mistral-nemo:latest',
+    name: 'Mistral NeMo (Ollama)',
+    description: 'Mistral NeMo 12B - via Ollama',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'codellama:latest',
+    name: 'Code Llama (Ollama)',
+    description: 'Meta Code Llama - Code generation, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 16384,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'deepseek-coder:latest',
+    name: 'DeepSeek Coder (Ollama)',
+    description: 'DeepSeek Coder - Code generation, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 16384,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'deepseek-r1:latest',
+    name: 'DeepSeek R1 (Ollama)',
+    description: 'DeepSeek R1 - Reasoning model, via Ollama',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'qwen2.5:latest',
+    name: 'Qwen 2.5 (Ollama)',
+    description: 'Alibaba Qwen 2.5 - General purpose, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'qwen2.5-coder:latest',
+    name: 'Qwen 2.5 Coder (Ollama)',
+    description: 'Qwen 2.5 Coder - Code generation, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'gemma2:latest',
+    name: 'Gemma 2 (Ollama)',
+    description: 'Google Gemma 2 - via Ollama',
+    maxTokens: 8192,
+    contextWindow: 8192,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'gemma3:latest',
+    name: 'Gemma 3 (Ollama)',
+    description: 'Google Gemma 3 - Latest, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'phi3:latest',
+    name: 'Phi-3 (Ollama)',
+    description: 'Microsoft Phi-3 Mini - Compact, via Ollama',
+    maxTokens: 4096,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'phi3.5:latest',
+    name: 'Phi-3.5 (Ollama)',
+    description: 'Microsoft Phi-3.5 - Latest compact, via Ollama',
+    maxTokens: 4096,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'cogito:latest',
+    name: 'Cogito (Ollama)',
+    description: 'Cogito - Reasoning model, via Ollama',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'minimax-m2.5:cloud',
+    name: 'Minimax M2.5 (Ollama Cloud)',
+    description: 'Minimax M2.5 - Cloud via Ollama',
+    maxTokens: 16384,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'ollama'
+  },
+  {
+    id: 'llama-3.3-70b-versatile',
+    name: 'Llama 3.3 70B (Groq)',
+    description: 'Llama 3.3 70B via Groq - ultra fast inference',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'llama-3.1-8b-instant',
+    name: 'Llama 3.1 8B (Groq)',
+    description: 'Llama 3.1 8B via Groq - ultra fast inference',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'llama-4-maverick-17b-128e-instruct',
+    name: 'Llama 4 Maverick (Groq)',
+    description: 'Llama 4 Maverick 17B via Groq',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'llama-4-scout-17b-16e-instruct',
+    name: 'Llama 4 Scout (Groq)',
+    description: 'Llama 4 Scout 17B via Groq',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'mixtral-8x7b-32768',
+    name: 'Mixtral 8x7B (Groq)',
+    description: 'Mixtral 8x7B via Groq - ultra fast inference',
+    maxTokens: 8192,
+    contextWindow: 32768,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'gemma2-9b-it',
+    name: 'Gemma 2 9B (Groq)',
+    description: 'Google Gemma 2 9B via Groq',
+    maxTokens: 8192,
+    contextWindow: 8192,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'qwen/qwen3-32b',
+    name: 'Qwen 3 32B (Groq)',
+    description: 'Alibaba Qwen 3 32B via Groq',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'moonshotai/kimi-k2-instruct',
+    name: 'Kimi K2 (Groq)',
+    description: 'Moonshot Kimi K2 Instruct via Groq',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'groq'
+  },
+  {
+    id: 'gemini-2.0-flash',
+    name: 'Gemini 2.0 Flash',
+    description: 'Google Gemini 2.0 Flash - 1M context (Google AI Studio)',
+    maxTokens: 8192,
+    contextWindow: 1000000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'google'
+  },
+  {
+    id: 'gemini-2.5-flash-preview-05-20',
+    name: 'Gemini 2.5 Flash Preview',
+    description: 'Google Gemini 2.5 Flash Preview (Google AI Studio)',
+    maxTokens: 8192,
+    contextWindow: 1000000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'google'
+  },
+  {
+    id: 'gemini-2.5-pro-preview-05-06',
+    name: 'Gemini 2.5 Pro Preview',
+    description: 'Google Gemini 2.5 Pro Preview (Google AI Studio)',
+    maxTokens: 8192,
+    contextWindow: 1000000,
+    supportsTools: true,
+    supportsVision: true,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'google'
+  },
+  {
+    id: 'llama-3.3-70b',
+    name: 'Llama 3.3 70B (Cerebras)',
+    description: 'Llama 3.3 70B via Cerebras - ultra fast',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'cerebras'
+  },
+  {
+    id: 'llama3.1-8b',
+    name: 'Llama 3.1 8B (Cerebras)',
+    description: 'Llama 3.1 8B via Cerebras - ultra fast',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'cerebras'
+  },
+  {
+    id: 'qwen-3-235b-a22b-instruct',
+    name: 'Qwen 3 235B (Cerebras)',
+    description: 'Qwen 3 235B A22B Instruct via Cerebras',
+    maxTokens: 8192,
+    contextWindow: 128000,
+    supportsTools: true,
+    supportsVision: false,
+    pricing: { input: 0, output: 0 },
+    isFree: true,
+    backend: 'cerebras'
+  }
+];
+
+export class FreeModelsProvider {
+  private client: OpenAI | null = null;
+  private config: ProviderConfig;
+  private activeBackend: string = 'openrouter';
+  private apiKeys: Record<string, string> = {};
+
+  constructor(config: ProviderConfig = {}) {
+    this.config = config;
+    this.loadApiKeys();
+    this.initializeClient();
+  }
+
+  private loadApiKeys(): void {
+    this.apiKeys = {
+      openrouter: this.config.apiKey || process.env.OPENROUTER_API_KEY || '',
+      ollama: process.env.OLLAMA_API_KEY || '',
+      nvidia: process.env.NVIDIA_API_KEY || '',
+      groq: process.env.GROQ_API_KEY || '',
+      google: process.env.GOOGLE_API_KEY || '',
+      cerebras: process.env.CEREBRAS_API_KEY || '',
+      cloudflare: process.env.CLOUDFLARE_API_KEY || '',
+    };
+  }
+
+  setApiKey(backend: string, apiKey: string): void {
+    this.apiKeys[backend] = apiKey;
+    if (this.activeBackend === backend) {
+      this.initializeClient();
+    }
+  }
+
+  getApiKey(backend: string): string {
+    return this.apiKeys[backend] || '';
+  }
+
+  private initializeClient(): void {
+    const backendConfig = FREE_MODEL_CONFIGS[this.activeBackend];
+    const baseUrl = this.config.baseUrl || backendConfig?.baseUrl || FREE_MODEL_CONFIGS.openrouter.baseUrl;
+    
+    let apiKey = this.config.apiKey || this.apiKeys[this.activeBackend] || '';
+
+    const headers: Record<string, string> = {
+      ...(this.config.customHeaders || {})
+    };
+
+    if (this.activeBackend === 'openrouter') {
+      headers['HTTP-Referer'] = 'https://codex-linux.app';
+      headers['X-Title'] = 'Codex Linux';
+    }
+
+    this.client = new OpenAI({
+      apiKey: apiKey || 'sk-dummy-key',
+      baseURL: baseUrl,
+      timeout: this.config.timeout || 60000,
+      maxRetries: this.config.maxRetries || 3,
+      defaultHeaders: headers
+    });
+  }
+
+  setBackend(backend: string, apiKey?: string): void {
+    const backendConfig = FREE_MODEL_CONFIGS[backend];
+    if (!backendConfig) {
+      throw new Error(`Unknown backend: ${backend}`);
+    }
+
+    this.activeBackend = backend;
+    if (apiKey) {
+      this.apiKeys[backend] = apiKey;
+    }
+    this.config.baseUrl = backendConfig.baseUrl;
+    this.config.apiKey = this.apiKeys[backend] || '';
+    this.initializeClient();
+  }
+
+  async sendMessage(
+    model: string,
+    messages: Array<{ role: string; content: string }>,
+    options?: {
+      signal?: AbortSignal;
+      onProgress?: (progress: number) => void;
+      extendedThinking?: boolean;
+      reasoningEffort?: 'low' | 'medium' | 'high';
+    }
+  ): Promise<{ content: string; metadata?: Record<string, any> }> {
+    if (!this.client) {
+      throw new Error('Client not initialized');
+    }
+
+    const openaiMessages = messages.map(m => ({
+      role: m.role as 'user' | 'assistant' | 'system',
+      content: m.content
+    }));
+
+    const requestParams: any = {
+      model,
+      messages: openaiMessages,
+      stream: false
+    };
+
+    const isNvidiaBackend = this.activeBackend === 'nvidia';
+    const supportsThinking = model.includes('kimi-k2') || model.includes('deepseek-r1');
+    
+    if (options?.extendedThinking) {
+      if (isNvidiaBackend && supportsThinking) {
+        requestParams.chat_template_kwargs = { thinking: true };
+      } else if (model.includes('deepseek-r1') || model.includes('qwq')) {
+        requestParams.reasoning_effort = options.reasoningEffort || 'medium';
+      }
+    }
+
+    if (isNvidiaBackend) {
+      requestParams.max_tokens = requestParams.max_tokens || 16384;
+      requestParams.temperature = 1.0;
+      requestParams.top_p = 1.0;
+    }
+
+    try {
+      const response = await this.client.chat.completions.create(requestParams, {
+        signal: options?.signal
+      });
+
+      const content = response.choices[0]?.message?.content || '';
+      const reasoning = (response.choices[0]?.message as any)?.reasoning_content;
+
+      return {
+        content,
+        metadata: {
+          model: response.model,
+          usage: response.usage,
+          finishReason: response.choices[0]?.finish_reason,
+          reasoning: reasoning || null,
+          extendedThinking: options?.extendedThinking || false,
+          reasoningEffort: options?.reasoningEffort,
+          backend: this.activeBackend
+        }
+      };
+    } catch (error: any) {
+      log.error('FreeModelsProvider sendMessage error:', error);
+      throw error;
+    }
+  }
+
+  async sendMessageStream(
+    model: string,
+    messages: Array<{ role: string; content: string }>,
+    callbacks?: {
+      onChunk?: (chunk: string) => void;
+      onComplete?: () => void;
+      onError?: (error: Error) => void;
+    }
+  ): Promise<{ content: string }> {
+    if (!this.client) {
+      throw new Error('Client not initialized');
+    }
+
+    const openaiMessages = messages.map(m => ({
+      role: m.role as 'user' | 'assistant' | 'system',
+      content: m.content
+    }));
+
+    let fullContent = '';
+
+    try {
+      const stream = await this.client.chat.completions.create({
+        model,
+        messages: openaiMessages,
+        stream: true
+      });
+
+      for await (const chunk of stream) {
+        const content = chunk.choices[0]?.delta?.content || '';
+        if (content) {
+          fullContent += content;
+          callbacks?.onChunk?.(content);
+        }
+      }
+
+      callbacks?.onComplete?.();
+      return { content: fullContent };
+    } catch (error: any) {
+      callbacks?.onError?.(error);
+      throw error;
+    }
+  }
+
+  listModels(): ProviderModel[] {
+    return FREE_MODELS;
+  }
+
+  async fetchOllamaModels(): Promise<ProviderModel[]> {
+    const ollamaUrl = this.config.baseUrl?.replace('/v1', '') || 'http://localhost:11434';
+    const headers: Record<string, string> = {};
+    if (this.apiKeys.ollama) {
+      headers['Authorization'] = `Bearer ${this.apiKeys.ollama}`;
+    }
+
+    try {
+      const response = await fetch(`${ollamaUrl}/api/tags`, {
+        headers,
+        signal: AbortSignal.timeout(5000)
+      });
+      
+      if (!response.ok) {
+        return [];
+      }
+
+      const data = await response.json() as { models: Array<{ name: string; details?: any }> };
+      
+      return (data.models || []).map(model => ({
+        id: model.name,
+        name: `${model.name} (Ollama Local)`,
+        description: `Local Ollama model: ${model.name}`,
+        maxTokens: 8192,
+        contextWindow: 32768,
+        supportsTools: true,
+        supportsVision: false,
+        pricing: { input: 0, output: 0 },
+        isFree: true,
+        backend: 'ollama'
+      }));
+    } catch (error) {
+      log.warn('Failed to fetch Ollama models:', error);
+      return [];
+    }
+  }
+
+  getBackends(): { id: string; name: string; requiresApiKey: boolean; description: string }[] {
+    return Object.entries(FREE_MODEL_CONFIGS).map(([id, config]) => ({
+      id,
+      name: config.provider,
+      requiresApiKey: config.requiresApiKey,
+      description: config.description
+    }));
+  }
+
+  async testConnection(): Promise<boolean> {
+    if (!this.client) {
+      return false;
+    }
+
+    const testModels: Record<string, string> = {
+      openrouter: 'meta-llama/llama-3.3-70b-instruct:free',
+      ollama: 'llama3.2:latest',
+      nvidia: 'meta/llama-3.3-70b-instruct',
+      groq: 'llama-3.1-8b-instant',
+      google: 'gemini-2.0-flash',
+      cerebras: 'llama-3.3-70b',
+      cloudflare: '@cf/meta/llama-3.1-8b-instruct'
+    };
+
+    try {
+      const testModel = testModels[this.activeBackend] || 'meta-llama/llama-3.3-70b-instruct:free';
+      const response = await this.client.chat.completions.create({
+        model: testModel,
+        messages: [{ role: 'user', content: 'test' }],
+        max_tokens: 1
+      });
+      return true;
+    } catch (error: any) {
+      log.warn('FreeModelsProvider connection test failed:', error.message);
+      return error.status !== 401;
+    }
+  }
+}
+
+export { FREE_MODELS, FREE_MODEL_CONFIGS };