Small LLM primitives for typed provider execution, structured responses, routing, prompts, context, sessions, and metrics.

Task-state, loop-detection, scratchpad, and verification harness for improving small-LLM reliability in Pi.

Turn OCR bounding boxes into descriptors a small LLM can read to classify the text inside each box

Shared LLM generation policy for DeepSeek Harness session-title providers

Adaptive LLM router with tier-based model selection and fallback support.

Provider-neutral LLM service interface for the DeepSeek Harness

Local SQLite memory engine for Oh My Pi agents

Self-learning LLM runtime — TurboQuant KV-cache (6-8x compression), SONA adaptive learning, FlashAttention, speculative decoding, GGUF inference

Use Custom Media Queries in CSS

Provider-routed LLM request retry policy for the DeepSeek Harness

A JavaScript interface for chroma

pi-ai-backed DeepSeek adapter for the DeepSeek Harness LLM seam (design-verification twin of dsh-llm-deepseek)

DeepSeek chat-completions adapter for the DeepSeek Harness LLM seam

Information on LLM models, context window token limit, output token limit, pricing and more

LLM eval & testing toolkit

Typescript bindings for langchain

Hardware accelerated language model chats on browsers

Additive request-field registry for the official DeepSeek LLM API adapter

Veryfront first-party extension package for ext-llm-openai

Display language model outputs in your React project.