{ "gemini": { "family": "Gemini", "provider": "gemini", "models": [ { "model": "gemini-2.5-flash-lite", "suffix": "2.5 Flash Lite", "short_suffix": "2.5FL", "description": "Fast and economical model", "thinking_modes": [ "none", "LOW", "MEDIUM", "HIGH" ], "grounding": true, "tools": true }, { "model": "gemini-3.1-flash-lite", "suffix": "3.1 Flash Lite", "short_suffix": "3.1FL", "description": "High speed, efficient thinking model", "thinking_modes": [ "none", "MINIMAL", "LOW", "MEDIUM", "HIGH" ], "grounding": true, "tools": true }, { "model": "gemini-3.5-flash-lite", "suffix": "3.5 Flash Lite", "short_suffix": "3.5FL", "description": "Balanced speed and quality", "thinking_modes": [ "none", "MINIMAL", "LOW", "MEDIUM", "HIGH" ], "grounding": true, "tools": true }, { "model": "gemini-3.5-flash", "suffix": "3.5 Flash", "short_suffix": "3.5F", "description": "High performance intermediate model", "thinking_modes": [ "none", "MINIMAL", "LOW", "MEDIUM", "HIGH" ], "grounding": true, "tools": true }, { "model": "gemini-3.7-flash", "suffix": "3.7 Flash", "short_suffix": "3.7F", "description": "Advanced reasoning and fast performance", "thinking_modes": [ "none", "LOW", "MEDIUM", "HIGH" ], "grounding": true, "tools": true }, { "model": "gemini-3.8-flash", "suffix": "3.8 Flash", "short_suffix": "3.8F", "description": "Next-gen workhorse model engineered for reasoning, coding, and autonomous agents", "thinking_modes": [ "none", "LOW", "MEDIUM", "HIGH" ], "grounding": true, "tools": true } ] }, "deepseek": { "family": "DeepSeek", "provider": "deepseek", "models": [ { "model": "deepseek-flash", "suffix": "4.1 Flash", "short_suffix": "4.1F", "description": "High-efficiency native reasoning model for fast, low-latency, and cost-effective generation", "thinking_modes": [ "none", "low", "high", "max" ], "grounding": true, "tools": true } ] }, "zai": { "family": "Z.ai", "provider": "zai", "models": [ { "model": "glm-5.3-flash", "suffix": "GLM 5.3 Flash", "short_suffix": "5.3F", "description": "Fast, low-cost model for high-volume tasks; thinking depth set via reasoning effort", "thinking_modes": [ "low", "high", "max" ], "grounding": true, "tools": true }, { "model": "glm-5.3", "suffix": "GLM 5.3", "short_suffix": "5.3", "description": "Flagship model for reasoning, coding, and agentic tasks; thinking depth set via reasoning effort", "thinking_modes": [ "low", "high", "max" ], "grounding": true, "tools": true } ] }, "anthropic": { "family": "Anthropic", "provider": "anthropic", "models": [ { "model": "claude-haiku-4-5", "suffix": "Haiku 4.5", "short_suffix": "H4.5", "description": "Fast, low-cost model for high-volume tasks and lightweight agents", "thinking_modes": [ "none", "low", "medium", "high" ], "grounding": true, "tools": true }, { "model": "claude-sonnet-5", "suffix": "Sonnet 5", "short_suffix": "S5", "description": "Balanced production model for general-purpose intelligence, coding, and tool use", "thinking_modes": [ "none", "low", "medium", "high" ], "grounding": true, "tools": true }, { "model": "claude-opus-5-5", "suffix": "Opus 5.5", "short_suffix": "O5.5", "description": "Frontier model for complex reasoning, coding, and agentic workflows", "thinking_modes": [ "none", "low", "medium", "high" ], "grounding": true, "tools": true } ] }, "openai": { "family": "OpenAI", "provider": "openai", "models": [ { "model": "gpt-5.4-nano", "suffix": "5.4 Nano", "short_suffix": "5.4N", "description": "Ultra-low-cost, high-speed model for lightweight tasks, classification, and extraction", "thinking_modes": [ "none", "low", "medium", "high", "xhigh" ], "grounding": true, "tools": true }, { "model": "gpt-5.4-mini", "suffix": "5.4 Mini", "short_suffix": "5.4M", "description": "High-throughput lightweight model for coding, subagent orchestration, and tool use", "thinking_modes": [ "none", "low", "medium", "high", "xhigh" ], "grounding": true, "tools": true }, { "model": "gpt-5.6-luna", "suffix": "5.6 Luna", "short_suffix": "5.6L", "description": "High-speed, cost-optimized lightweight model for high-volume tasks and fast routing", "thinking_modes": [ "none", "low", "medium", "high", "xhigh", "max" ], "grounding": true, "tools": true }, { "model": "gpt-5.6-terra", "suffix": "5.6 Terra", "short_suffix": "5.6T", "description": "Balanced frontier model for general-purpose intelligence, coding, and production workloads", "thinking_modes": [ "none", "low", "medium", "high", "xhigh", "max" ], "grounding": true, "tools": true }, { "model": "gpt-5.6-sol", "suffix": "5.6 Sol", "short_suffix": "5.6S", "description": "Large reasoning model for complex mathematics, coding, and autonomous workflows", "thinking_modes": [ "none", "low", "medium", "high", "xhigh", "max" ], "grounding": true, "tools": true }, { "model": "gpt-6-astra", "suffix": "6 Astra", "short_suffix": "6A", "description": "Flagship reasoning model for complex mathematics, deep coding, and autonomous multi-agent workflows", "thinking_modes": [ "low", "medium", "high", "xhigh", "max" ], "grounding": true, "tools": true } ] }, "mistral": { "family": "Mistral", "provider": "mistral", "experimental": true, "models": [ { "model": "ministral-3b-latest", "suffix": "3 3B", "short_suffix": "3-3B", "description": "Fast and compact edge model for lightweight, low-latency text tasks", "thinking_modes": [ null ], "grounding": false, "tools": true }, { "model": "ministral-14b-latest", "suffix": "3 14B", "short_suffix": "3-14B", "description": "Dense edge model delivering high-performance reasoning and vision", "thinking_modes": [ null ], "grounding": false, "tools": true }, { "model": "codestral-latest", "suffix": "Codestral", "short_suffix": "Code", "description": "State-of-the-art model specialized in code generation and development tasks", "thinking_modes": [ null ], "grounding": false, "tools": true }, { "model": "open-mistral-nemo", "suffix": "NeMo 12B", "short_suffix": "NeMo", "description": "12B multilingual model with state-of-the-art reasoning performance", "thinking_modes": [ null ], "grounding": false, "tools": true }, { "model": "zai-glm-5-3", "suffix": "GLM 5.3", "short_suffix": "GLM-5.3", "description": "Frontier third-party model from Z.ai hosted by Mistral for long-context reasoning, coding, and agent workflows", "thinking_modes": [ "low", "high", "max" ], "grounding": false, "tools": true }, { "model": "mistral-small-latest", "suffix": "Small 4", "short_suffix": "S4", "description": "Cost-efficient hybrid model for reasoning, coding, and general tasks", "thinking_modes": [ "none", "high" ], "grounding": false, "tools": true }, { "model": "mistral-medium-latest", "suffix": "Medium 3.5", "short_suffix": "M3.5", "description": "Frontier multimodal model tuned for agentic workflows and coding", "thinking_modes": [ "none", "high" ], "grounding": false, "tools": true }, { "model": "mistral-large-latest", "suffix": "3 Large", "short_suffix": "3L", "description": "Flagship frontier model for complex reasoning and large contexts", "thinking_modes": [ null ], "grounding": false, "tools": true } ] } }