CoFounder Model List

models.json

{
  "chat_models": {
    "gpt-4.1": {
      "display_name": "GPT-4.1",
      "description": "The top GPT-4.1 model, built for complex tasks. Excellent at following detailed instructions, advanced coding, and understanding very long documents.",
      "litellm_name": "azure/gpt-4.1",
      "cli": "opencode"
    },
    "claude-opus-4-6": {
      "display_name": "Azure Claude Opus 4.6",
      "description": "Anthropic's latest and most capable model. Superior coding, deep reasoning, and complex multi-step workflows. The best choice for the hardest engineering and planning tasks.",
      "litellm_name": "azure_ai/claude-opus-4-6",
      "cli": "claude"
    },
    "claude-sonnet-4-5": {
      "display_name": "Azure Claude 4.5 Sonnet",
      "description": "A significant upgrade to Claude Sonnet 4, delivering superior coding and reasoning while responding more precisely to instructions. It balances performance and efficiency for various use cases.",
      "litellm_name": "azure_ai/claude-sonnet-4-5",
      "cli": "claude"
    },
    "claude-opus-4-5": {
      "display_name": "Azure Claude Opus 4.5",
      "description": "Anthropic's previous-gen flagship. Still excellent for expert-level software engineering, complex planning, and automating multi-step workflows.",
      "litellm_name": "azure_ai/claude-opus-4-5",
      "cli": "claude"
    },
    "claude-haiku-4-5": {
      "display_name": "Azure Claude 4.5 Haiku",
      "description": "A faster, lighter Claude model with improved reasoning and efficiency\u2014ideal for quick, high-quality responses across coding, writing, and analysis tasks.",
      "litellm_name": "azure_ai/claude-haiku-4-5",
      "cli": "claude"
    },
    "gpt-5.4-pro": {
      "display_name": "GPT-5.4 Pro",
      "description": "OpenAI's most powerful model (March 2026). Best-in-class for complex reasoning, coding, math, and multi-step planning. Ideal for expert-level tasks that demand the highest accuracy and reliability.",
      "litellm_name": "azure/gpt-5.4-pro",
      "cli": "codex"
    },
    "gpt-5.4": {
      "display_name": "GPT-5.4",
      "description": "The latest GPT-5 series model (March 2026). Excellent balance of speed and intelligence for coding, analysis, and creative tasks. A strong default for most use cases.",
      "litellm_name": "azure/gpt-5.4",
      "cli": "codex"
    },
    "gpt-5.3-codex": {
      "display_name": "GPT-5.3 Codex",
      "description": "OpenAI's latest code-specialized model. Optimized for autonomous coding tasks, tool use, and software engineering workflows with full-auto approval mode support.",
      "litellm_name": "azure/gpt-5.3-codex",
      "cli": "codex"
    },
    "gpt-5.2": {
      "display_name": "GPT-5.2",
      "description": "GPT-5.2 powers fast, context-aware chat experiences with adaptive reasoning and robust safety guardrails. Strong all-rounder for interactive coding and analysis.",
      "litellm_name": "azure/gpt-5.2",
      "cli": "codex"
    },
    "gpt-5-pro": {
      "display_name": "GPT-5 Pro",
      "description": "OpenAI's extended-thinking model for the most demanding tasks. Uses deeper reasoning for superior accuracy on complex math, science, and coding problems.",
      "litellm_name": "azure/gpt-5-pro",
      "cli": "codex"
    },
    "gpt-5.1": {
      "display_name": "GPT-5.1",
      "description": "A solid mid-generation GPT-5 model with strong reasoning, coding, and instruction-following capabilities. Good balance of cost and performance.",
      "litellm_name": "azure/gpt-5.1",
      "cli": "codex"
    },
    "o4-mini": {
      "display_name": "o4-mini",
      "description": "OpenAI's latest compact reasoning model. Fast and cost-effective with strong performance on coding, math, and logic tasks. Great for quick iterations.",
      "litellm_name": "azure/o4-mini",
      "cli": "codex"
    },
    "o3": {
      "display_name": "o3",
      "description": "OpenAI's advanced reasoning model with chain-of-thought capabilities. Excels at complex multi-step problems requiring deep logical analysis.",
      "litellm_name": "azure/o3",
      "cli": "codex"
    },
    "o3-pro": {
      "display_name": "o3 Pro",
      "description": "The most powerful reasoning model from OpenAI. Extended thinking time for the hardest problems in math, science, and engineering. Use when accuracy matters more than speed.",
      "litellm_name": "azure/o3-pro",
      "cli": "codex"
    },
    "grok-3": {
      "display_name": "Grok 3",
      "description": "xAI's powerful model with strong reasoning and real-time knowledge. Handles text and images with a large context window. Good for research and analysis tasks.",
      "litellm_name": "azure/grok-3",
      "cli": "opencode"
    },
    "grok-3-mini": {
      "display_name": "Grok 3 Mini",
      "description": "A fast, cost-effective xAI model. Good reasoning capabilities at lower latency. Ideal for quick tasks and interactive chat.",
      "litellm_name": "azure/grok-3-mini",
      "cli": "opencode"
    },
    "DeepSeek-V3.2": {
      "display_name": "DeepSeek V3.2",
      "description": "A powerful open-source model running on Azure. Top competitor in reasoning, coding, and reliability. Excellent cost-performance ratio for demanding tasks.",
      "litellm_name": "azure/DeepSeek-V3.2",
      "cli": "opencode"
    },
    "Kimi-K2.5": {
      "display_name": "Kimi K2.5",
      "description": "MoonshotAI's latest model running on Azure. A powerful Mixture-of-Experts model designed for complex agentic tasks, coding, and reasoning with excellent multilingual support.",
      "litellm_name": "azure/Kimi-K2.5",
      "cli": "opencode"
    },
    "openrouter/google/gemini-3.1-pro-preview": {
      "display_name": "Gemini 3.1 Pro Preview",
      "description": "Google's latest and most advanced AI model. State-of-the-art for complex reasoning, coding, math, and science. Uses enhanced thinking capabilities for superior accuracy on the hardest problems.",
      "litellm_name": "gemini/gemini-3.1-pro-preview",
      "cli": "gemini"
    },
    "openrouter/google/gemini-3.1-pro-preview": {
      "display_name": "Gemini 3.1 Flash Lite Preview",
      "description": "Google's fastest and most cost-effective model. Excellent for high-volume tasks, quick coding, and real-time interactions. Great speed-to-quality ratio.",
      "litellm_name": "gemini/gemini-3.1-flash-lite-preview",
      "cli": "gemini"
    },
    "openrouter/anthropic/claude-sonnet-4.5": {
      "display_name": "Claude 4.5 Sonnet",
      "description": "A significant upgrade to Claude Sonnet 4, delivering superior coding and reasoning while responding more precisely to instructions. Balances performance and efficiency.",
      "litellm_name": "openrouter/anthropic/claude-sonnet-4.5",
      "cli": "claude"
    },
    "openrouter/anthropic/claude-haiku-4.5": {
      "display_name": "Claude 4.5 Haiku",
      "description": "A faster, lighter Claude model with improved reasoning and efficiency\u2014ideal for quick, high-quality responses across coding, writing, and analysis tasks.",
      "litellm_name": "openrouter/anthropic/claude-haiku-4.5",
      "cli": "claude"
    },
    "openrouter/anthropic/claude-opus-4.5": {
      "display_name": "Claude Opus 4.5",
      "description": "Anthropic's previous-gen flagship via OpenRouter. Expert-level software engineering, complex planning, and multi-step workflows.",
      "litellm_name": "openrouter/anthropic/claude-opus-4.5",
      "cli": "claude"
    },
    "openrouter/moonshotai/kimi-k2": {
      "display_name": "MoonshotAI: Kimi K2",
      "description": "Kimi K2 from Moonshot AI is a powerful, open-source 1-trillion-parameter Mixture-of-Experts model designed for complex agentic tasks, coding, and reasoning.",
      "litellm_name": "moonshot/kimi-k2-0711-preview",
      "cli": "opencode"
    },
    "openrouter/qwen/qwen3-vl-235b-a22b-thinking": {
      "display_name": "Qwen3-VL Thinking",
      "description": "A powerful model that can \"see\" and \"think.\" It understands images, video, and text to solve complex math and science problems. Can turn sketches into working code and handle multilingual documents.",
      "litellm_name": "openrouter/qwen/qwen3-vl-235b-a22b-thinking",
      "cli": "opencode"
    },
    "openrouter/qwen/qwen3-max": {
      "display_name": "Qwen3 Max",
      "description": "A top-tier model from Qwen. Great for coding, science, and detailed instructions. Speaks over 100 languages with strong translation and high-quality answers.",
      "litellm_name": "openrouter/qwen/qwen3-max",
      "cli": "opencode"
    },
    "openrouter/qwen/qwen3-coder-plus": {
      "display_name": "Qwen3 Coder Plus",
      "description": "A powerful AI specifically built for programming. Acts as an autonomous coding assistant, capable of writing code, using tools, and interacting with environments.",
      "litellm_name": "openrouter/qwen/qwen3-coder-plus",
      "cli": "opencode"
    },
    "openrouter/z-ai/glm-4.7": {
      "display_name": "GLM-4.7",
      "description": "Excels at understanding long, complex documents and images. Can process massive amounts of information and turn screenshots into website code.",
      "litellm_name": "openrouter/z-ai/glm-4.7",
      "cli": "opencode"
    },
    "openrouter/mistralai/mistral-large-2512": {
      "display_name": "Mistral Large 3 2512",
      "description": "Mistral's most powerful model, built for top performance on demanding tasks with a highly efficient architecture.",
      "litellm_name": "openrouter/mistralai/mistral-large-2512",
      "cli": "opencode"
    },
    "openrouter/deepseek/deepseek-v3.2-speciale": {
      "display_name": "DeepSeek V3.2 Speciale",
      "description": "A super-charged DeepSeek V3.2 for maximum thinking and complex automation. Top competitor in reasoning, coding, and reliability.",
      "litellm_name": "openrouter/deepseek/deepseek-v3.2-speciale",
      "cli": "opencode"
    },
    "openrouter/x-ai/grok-4": {
      "display_name": "Grok 4",
      "description": "xAI's latest powerful model for deep thinking and complex tasks. Handles text and images with tools and 256k token context.",
      "litellm_name": "openrouter/x-ai/grok-4",
      "cli": "codex"
    },
    "openrouter/x-ai/grok-4.1-fast": {
      "display_name": "Grok 4.1 Fast",
      "description": "xAI's fastest model for automation and real-world tasks. Excellent for customer support and research with 2M context for interactive workflows.",
      "litellm_name": "openrouter/x-ai/grok-4.1-fast",
      "cli": "codex"
    }
  }
}