Frontier Models,
One Workspace

OpenAI's flagship open-weight model with 120B parameters, served on Groq's ultra-fast hardware. Delivers GPT-4+ class reasoning and writing with strong performance across coding, analysis, and conversation.

Google's most capable Flash model — see images, read PDFs, analyze documents, and hold complex conversations at remarkable speed. The go-to model when you need fast, smart, and versatile in one.

NVIDIA's frontier reasoning model — 550B parameters with only 55B active per pass (MoE). Built for long-running agentic workflows, deep research, and complex multi-step planning. 262K token context with extended thinking. Ranked #1 in speed among frontier models at 420+ tok/s.

xAI's flagship model for low-hallucination answers, agentic tool calling, and instruction following. Supports text and image input, structured outputs, function calling, configurable reasoning, and a 1M token context window for long-document analysis and complex workflows.

DeepSeek's latest flagship with 1M token context and up to 384K output tokens. Supports thinking and non-thinking modes, tool calls, and JSON output. Excellent for coding, analysis, and long-context tasks. Pro users can upgrade to V4 Pro mid-chat via the ☄️ toolbar button — a Pro-only feature.

Anthropic's best balance of speed and intelligence. Extended thinking + adaptive thinking for complex reasoning. 1M token context window. Excels at coding, writing, analysis, and nuanced tasks. Top-tier quality for critical work.

Generates high-quality images from text using the FLUX model. Creates 1024×1024 images. Works best with detailed, descriptive prompts. Fast generation, unlimited use.