🔥 Current Recommended Models Below are the currently stable and popular models. For the complete model list and real-time pricing, please visit MixRoute Model Marketplace .
When calling API , please ensure the AI model name matches the naming convention in the MixRoute Model Marketplace ; otherwise, the request will fail to execute. Model Categories 🤖 OpenAI Series Reasoning Models Model Name Model ID Features Recommended For o4-mini o4-mini Lightweight reasoning model Programming tasks o3-mini o3-mini Latest reasoning model, excellent value Complex reasoning, math, coding
GPT Series Model Name Model ID Context Features Recommended For GPT 5.5 ⭐ gpt-5.5 1M Next-gen intelligent flagship, native computer operation, 82.7% Terminal-Bench 2.0 score Conversation, reasoning, code capability, agent tasks GPT 5.4 Pro gpt-5.4-pro 1M Professional version, native computer operation, tool search capability, most efficient reasoning model Conversation, code capability, copywriting GPT 5.4 gpt-5.4 1M First general-purpose native computer operation model, supports large tool ecosystems Conversation, copywriting GPT 5.4 Mini gpt-5.4-mini 400K Lightweight efficient version, maintains powerful reasoning while optimizing cost Conversation, reasoning, copywriting, code capability GPT 5.3 Codex gpt-5.3-codex 272K 25% faster than predecessor, best SWE-Bench Pro and Terminal-Bench performance, supports interactive guidance Conversation, code capability, copywriting
Image Generation Models Model Name Model ID Features GPT-Image-2 ⭐ gpt-image-2 Latest flagship image generation model GPT-Image-1.5 gpt-image-1.5 Upgraded image generation, stronger understanding GPT-Image-1 gpt-image-1 OpenAI official image generation
🎭 Claude Series (Anthropic) Claude 4 Series (Latest) Model Name Model ID Context Features Recommended For Claude Opus 4.8 ⭐ claude-opus-4-8 1M Latest flagship, hybrid reasoning with adaptive thinking, production-grade code capability. Conversation, reasoning, code capability, agent tasks Claude Opus 4.7 claude-opus-4-7 1M Set new standards for production code, complex agents, and complex office tasks Conversation, reasoning, code capability, copywriting Claude Sonnet 4.6 ⭐ claude-sonnet-4-6 200K Frontier performance, 70% more token-efficient than Sonnet 4.5, zero hallucinated links, suitable for customer agents and browser automation Conversation, code capability, copywriting Claude Sonnet 4.5 claude-sonnet-4-5-20250929 200K Agent capability progress, enhanced tool handling, memory management, and context processing performance Conversation, reasoning, copywriting, code capability Claude Haiku 4.5 claude-haiku-4-5-20251001 200K Anthropic’s most efficient model, fast and cost-effective near-frontier performance Conversation, reasoning, copywriting, code capability
🌟 Google Gemini Series Model Name Model ID Context Features Recommended For Gemini 3.5 Flash ⭐ gemini-3.5-flash 1M Latest fast version, optimized for speed-sensitive applications, cost-effective Conversation, reasoning, copywriting, code capability Gemini 3.1 Pro Preview gemini-3.1-pro-preview 1M Complex reasoning and analysis capability, latest flagship preview Conversation, reasoning, copywriting, code capability Gemini 3.1 Flash Lite Preview gemini-3.1-flash-lite-preview 1M Ultra-fast, low-cost, suitable for simple tasks and classification Conversation, copywriting Gemini 3.1 Flash Image Preview gemini-3.1-flash-image-preview 1M Image generation model, brings reasoning capability to image generation Text-to-image, image-to-image, image generation
🚀 xAI Grok Series Model Name Model ID Context Features Recommended For Grok 4.20 Reasoning ⭐ grok-4.20-beta-0309-reasoning 2M Latest flagship reasoning model, real-time knowledge integration, more open content policy Conversation, reasoning, copywriting, code capability Grok Code Fast grok-code-fast-1 256K Code-specific fast model, optimized for programming tasks Conversation, code capability Grok-4 grok-4 256K Powerful flagship model, unparalleled performance in natural language, mathematics, and reasoning Conversation, reasoning, copywriting, code capability Grok-4 Fast Reasoning grok-4-fast-reasoning 2M Cost-effective, fast reasoning version Conversation, reasoning, copywriting, code capability Grok-3 grok-3 128K Excels in enterprise applications, professional knowledge in finance, healthcare, law, and science Conversation, copywriting
🔍 DeepSeek Series Model Name Model ID Context Features Recommended For DeepSeek V4 Pro ⭐ deepseek-v4-pro 1M DeepSeek’s latest flagship model, top-tier reasoning capability, enhanced Agent capability Conversation, reasoning, copywriting, code capability DeepSeek V4 Flash deepseek-v4-flash 1M V4 fast version, maintains powerful reasoning while optimizing response speed Conversation, reasoning, copywriting, code capability DeepSeek V3.2 deepseek-v3-2-251201 164K V3 upgraded version, improved overall capability, cost-effective General scenario DeepSeek V3.1 250821 deepseek-v3.1-250821 128K Hybrid reasoning model, supports thinking and non-thinking modes, higher thinking efficiency than DeepSeek-R1 Conversation, reasoning, copywriting, code capability
🐘 Chinese Model Series Zhipu AI (GLM) Model Name Model ID Context Features Recommended For glm-4-plus glm-4-plus 128K Zhipu’s flagship model with strong coding and multi-step reasoning capabilities Chat, thinking, coding glm-4.5-air glm-4.5-air 128K High cost-performance version, suitable for fast response scenarios Chat, copywriting
Alibaba Qwen Model Name Model ID Context Features Recommended For Qwen Image Plus qwen-image-plus 32K 20B params image generation, excellent text rendering Text-to-image, image generation Qwen 3.7 Max qwen3.7-max 128K Alibaba’s latest flagship model, comprehensive capability improvements Chat, thinking, coding, copywriting
Moonshot Kimi Series Model Name Model ID Context Features Recommended For Kimi K2.6 ⭐ kimi-k2.6 200K Moonshot’s latest flagship model, powerful Chinese understanding and document processing capability Conversation, reasoning, copywriting, code capability kimi-k2.5 ⭐ kimi-k2.5 256K Moonshot’s most versatile model, native multimodal architecture, supports visual and text input, thinking and non-thinking modes Conversation, reasoning, copywriting, code capability Kimi K2 250905 kimi-k2-250905 256K MoE architecture foundation model with strong code and Agent capability, 1T total parameters, 32B active parameters Conversation, reasoning
Seedance Model Name Model ID Features Seedance 2.0 ⭐ dreamina-seedance-2-0-260128 Dreamina’s latest flagship video gen Seedance 2.0 Fast dreamina-seedance-2-0-fast-260128 Fast version, optimized speed Seedance 1.5 Pro seedance-1-5-pro-251215 High-quality professional video gen Seedance 1.0 Pro seedance-1-0-pro-250528 Base video generation model Seedance 1.0 Pro Fast seedance-1-0-pro-fast-251015 Fast version, lightweight tasks
🛠️ Usage Recommendations Model Selection Guide
Programming Top Performance : Claude Opus 4.7, GPT 5.5, Claude Sonnet 4.6, GPT 5.3 CodexCost-effective : Claude Haiku 4.5, GPT 5.4 Mini, DeepSeek V4 Flash, Grok-4 Fast ReasoningAlternatives : o3, Gemini 3.1 Pro Preview, qwen3.7-max, kimi-k2.6
Content Writing Recommended : GPT 5.5, Claude Opus 4.7, Claude Sonnet 4.6, glm-4-plusAlternatives : Gemini 3.1 Pro Preview, Kimi K2.6, glm-4.5-air, DeepSeek V4 Pro
Fast Response Recommended : Claude Haiku 4.5, Gemini 3.5 Flash, Grok 4.20 ReasoningAlternatives : glm-4.5-air, GPT 5.4 Mini, DeepSeek V4 Flash
Image Generation Recommended : GPT-Image-2, Gemini 3.1 Flash Image Preview, Qwen Image Plus
Long Context Processing
Ultra-long Context : Gemini 3.1 Pro Preview (1M), Gemini 3.5 Flash (1M), Claude Opus 4.7 (1M), GPT 5.5 (1M)Coding Scenarios : Claude 4 Series, kimi-k2.6, Kimi K2 Thinking, GPT 5.3 Codex
Cost Optimization Suggestions
Tiered Usage : Use cheaper models for simple tasks, advanced models for complex tasksTesting and Optimization : Test with small models first, scale to larger models once requirements are clearBatch Processing : Use Mini versions for a large number of similar tasksCache Reuse : Cache results for repetitive queries
API Documentation Detailed API reference
Quick Start Integration guide
Model list is continuously updated. We add newly released excellent models promptly. Contact support for specific models or bulk needs.
