LLM Minimal

Smallest streaming-LLM example — TemplateStream and EventStream side by side

Page actions AI-ready formats and sharing
Open LLM text
Share with AI
Ask Claude Ask ChatGPT Ask Gemini Ask Copilot

Overview

LLM Minimal is the smallest path to streaming tokens in Chirp — two forms on one page, simulated tokens by default (no Ollama or API keys required). Reach for it when Streaming answers describes the transport × client matrix and you want a runnable reference.

Location:examples/standalone/llm_minimal/

What it demonstrates

Form Transport Client wiring
First form (full page) TemplateStream Plain<form method="post"> — no hx-target
Second form (in place) EventStream htmx swaps in an SSE panel with parametricsse-connect

Chirp'stemplate_stream_client_shapecontract warns if you htmx-swap a TemplateStreamresponse into a div. See Contract categories for the streaming-related checks.

Run

PYTHONPATH=src python examples/standalone/llm_minimal/app.py

Open http://localhost:8000/ and try both forms.