Latest LLM Models

State-of-the-art general-purpose model focused on long-horizon instruction following, reasoning, and coding at a lower cost tier.

OpenAI's latest flagship model for agentic work, advanced coding, research, and professional workflows.

Moonshot AI's latest frontier model with strong long-context reasoning, agentic coding, and multimodal capabilities.

Anthropic's most capable generally available model for complex reasoning, agentic coding, long-running tasks, and stronger vision.

Smaller GPT-5.4 variant for fast, efficient coding and professional tasks with lower cost and latency.

The lightest GPT-5.4 variant, designed for high-volume, low-latency tasks while keeping the GPT-5.4 family behavior.

Compact multimodal model built for general chat, coding, agentic tasks, and complex reasoning.

xAI's multi-agent Grok model for collaborative reasoning, tool use, and advanced long-context workflows.

Latest mainstream Grok release for chat, reasoning, and tool-assisted workflows in xAI's platform.

OpenAI's frontier model for complex professional work, combining strong reasoning, coding, and large-context tool use.

Google's flagship Gemini model for complex multimodal tasks, creative work, coding, and advanced reasoning.

Anthropic's balanced model focused on strong reasoning, coding, and speed for day-to-day production workloads.

Earlier Kimi release focused on improved reasoning and multimodal capabilities before the K2.6 update.

Google's fast Gemini model optimized for responsive multimodal workloads and agentic coding at scale.

Mistral AI's flagship open-weight multimodal and multilingual model with strong agentic capabilities and a 256k context window.

Faster Grok variant for lower-latency inference, large-context workloads, and enterprise API usage.

Anthropic's fastest Claude 4.5 model, built for high-throughput tasks where latency and cost matter most.

Meta's open-weight multimodal model for general AI workloads, coding, and efficient large-scale deployment.

Meta's long-context open-weight multimodal model optimized for deep analysis across massive documents and codebases.