Chat with token compression built in

Pick an LLM above and start typing. ContextForge auto-compresses long context before it's sent — so big prompts cost fewer tokens.

Using demo key — limited free testing. Stateless — nothing is saved. v1.