Skip to content

Instantly share code, notes, and snippets.

@thomas-byern
Created February 21, 2026 18:19
Show Gist options
  • Select an option

  • Save thomas-byern/2823b5d3923789f720d3c34a5db6dfdd to your computer and use it in GitHub Desktop.

Select an option

Save thomas-byern/2823b5d3923789f720d3c34a5db6dfdd to your computer and use it in GitHub Desktop.
Local vs. Cloud LLMs
Factor Local LLMs Cloud LLMs
Privacy Data stays on your network Trust the provider
Capability ceiling Limited by your hardware State-of-the-art available
Latency (small tasks) Often faster, no network Network round-trip
Latency (complex tasks) Often much slower Optimized infrastructure
Cost model High fixed, low marginal Low fixed, scales with use
Maintenance Your responsibility Provider's responsibility
Context window (practical) Usually 4K-32K usable 100K+ available
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment