Текстовые AI-модели

Универсальная языковая модель для диалога и повседневных задач.

Универсальная языковая модель для диалога и повседневных задач.

Fast, low-cost multimodal Qwen model for high-volume chat and tool use.

High-quality long-context Qwen model for complex reasoning and agents.

General-purpose multimodal Qwen model with long context and tool use.

Qwen flagship long-context model for complex reasoning and agents.

Multimodal Qwen model for chat, vision, and tool use.

Fast multimodal Qwen model for low-cost chat and tool use.

Qwen3.8 Max for high-capability multimodal reasoning and tools.

Qwen3.8 Max Prime premium tier for the hardest reasoning and agents.

Fast omni-modal Qwen model for low-cost chat and tool use.

Anthropic frontier model for complex reasoning, coding, and long-running agents.

Anthropic flagship for frontier reasoning, coding, and long-running agents.

Sharp reasoning with fast responses — a strong Anthropic default for most work.

Rolling alias for the current DeepSeek fast tier (pass-through on OpenRouter).

Rolling alias for the current DeepSeek pro tier (pass-through on OpenRouter).

Универсальная языковая модель для диалога и повседневных задач.

Универсальная языковая модель для диалога и повседневных задач.

Универсальная языковая модель для диалога и повседневных задач.

Универсальная языковая модель для диалога и повседневных задач.

Fast, inexpensive Gemini model for high-volume multimodal chat.

Previous-generation Gemini Pro — the cheapest way to get Pro-tier multimodal reasoning.

Gemini 3.1 Pro preview for high-capability multimodal reasoning and tool use.

Fast, inexpensive Gemini 3.5 model for high-volume multimodal chat.

Fast multimodal Gemini 3.8 model for chat, vision, audio, and tool use.

Универсальная языковая модель для диалога и повседневных задач.

Meta Llama 3.1 8B instruction-tuned chat model served by Segmind.

Agentic coding and multimodal model with a 1M-token context window.

General-purpose Kimi model for chat, vision, and tool use.

Kimi coding model tuned for programming, debugging, and agentic tool use.

Moonshot AI flagship long-context model for complex reasoning, vision, and agents.

GPT-5.6 Luna for efficient text chat and tool use.

GPT-5.6 Sol for high-capability text chat and reasoning.

GPT-5.6 Terra for high-capability text chat and reasoning.

Fast, low-cost GPT-5 model for lightweight chat and tool use.

GPT-6 Astra flagship for high-capability multimodal chat and reasoning.

GPT-6 Astra Pro flagship with extended compute for hard reasoning.

GPT-6 Luna for efficient text chat and tool use.

GPT-6 Luna Pro for efficient text chat with extended compute.

GPT-6 Sol for high-capability text chat and reasoning.

GPT-6 Sol Pro for high-capability reasoning with extended compute.

Rolling alias for the current OpenAI Astra flagship (pass-through on OpenRouter).

Rolling alias for the current OpenAI Luna tier (pass-through on OpenRouter).

Large open-weight reasoning and tool-use model.

Compact open-weight reasoning and tool-use model.

Rolling alias for the current OpenAI Sol flagship (pass-through on OpenRouter).

Rolling alias for the current OpenAI Terra tier (pass-through on OpenRouter).

xAI flagship model for reasoning, vision, and tool use.

Rolling alias for the current xAI Grok flagship (pass-through on OpenRouter).

Fast Xiaomi MiMo model for low-cost chat and tool use.

Xiaomi MiMo Pro for capable multimodal reasoning and tools.

Xiaomi MiMo Pro tier served on ultraspeed inference for low-latency reasoning.

Z.AI flagship coding, reasoning, and tool-use model.

Z.ai flagship model for coding, reasoning, and tool use.

Fast Z.ai model for low-latency chat and tool use.

Fast Z.ai GLM 5.3 tier with vision for low-latency chat and tool use.

Z.ai premium GLM 5.3 tier for the hardest reasoning and agentic tasks.

Rolling alias for the current Z.AI GLM fast tier (pass-through on OpenRouter).

Rolling alias for the current Z.AI GLM flagship (pass-through on OpenRouter).