Context Explorer

Understand context windows, memory usage, and model fit

Context Explorer helps you understand how model context size, VRAM, concurrency, and engine behavior interact before you run real workloads.

Open it from Model Hub > Context Explorer.

Major Parts

Guided modes

Start with guided paths like Fit, Serve, and Learn, or jump directly into Free Explorer.

Interactive lessons

Use short lessons to learn context budgeting, overflow behavior, and concurrency tradeoffs.

Engine behavior simulation

Compare behavior across Ollama, vLLM, MLX, and llama.cpp, including overflow handling and context controls.

Scenario presets

Load presets, adjust context length and concurrency, then simulate to see memory impact and fit status.

Using Context Explorer

  1. Pick an engine, model size, and hardware profile.
  2. Adjust context length, precision, and concurrency.
  3. Run simulations and inspect memory buckets.
  4. Check fit status before applying similar settings in production.

Image placeholder: Context Explorer showing memory bucket bar and engine controls

Tip

Use Context Explorer when tuning local models for multi-user or long-context workflows.