Laya · Fast Decisions & Sorting

Input

State — text or JSON to decide about Questions (choice = pick one label · score = ordinal level · noul = yes/no probability)

Result

Pick a preset or write your own, then press Decide.

Sort items into categories

Items — one per line Categories — one per line, as name: description Question (optional) Shared context (optional)

Each item is classified independently.

Result

Results grouped by category appear here.

Rank items by a criterion

Items — one per line Criterion (a yes/no property) Shared context (optional)

Items are ordered by P(yes).

Result

Ranked list appears here.

LLM chat local · CPU · streaming

Model ⚡ Disable thinking

⚡ Stream demos (long output, watch tokens arrive)

⏳ Await demos (short answer, returned in one piece)

System prompt (optional) Your message

Max tokens Temperature keep conversation

Uses the API key saved in this browser (🔑 top right). Stream shows tokens live (SSE). Await waits for the full reply. CPU only: Gemma 2B ~18 tok/s, 8K-token context; one request at a time (others queue). Thinking toggles only apply to reasoning models — Gemma 2B answers directly, so they are greyed out. Prefer Open WebUI pointed at https://theetay.com/v1.

Reply 💭 Show thinking

statusidlefirst token–tokens/s–tokens–total–

Pick a preset or type a message, choose Stream or Await, then Send.

⚡ Laya

Input

Result

Sort items into categories

Result

Rank items by a criterion

Result

LLM chat local · CPU · streaming

Reply 💭 Show thinking

🔑 API key required