DEV Community

#llm

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
DeepSeek Harness got append-only right. Its token projection still misses what compaction costs.

DeepSeek Harness got append-only right. Its token projection still misses what compaction costs.

1
Comments 4
17 min read
Build 1-Click LLM Chatbot

Build 1-Click LLM Chatbot

Comments
5 min read
RAG na Prática: Como Avaliar Retrieval Sem Virar "Achei Que Ficou Bom"

RAG na Prática: Como Avaliar Retrieval Sem Virar "Achei Que Ficou Bom"

1
Comments
5 min read
Don't trust "Done." — forcing AI agents to re-fetch reality before they report completion

Don't trust "Done." — forcing AI agents to re-fetch reality before they report completion

1
Comments 10
7 min read
Agentic Loops: Quick Guide

Agentic Loops: Quick Guide

Comments
2 min read
How Much Does KV Cache Actually Save? A Real Look at Chinese LLM Cache Pricing

How Much Does KV Cache Actually Save? A Real Look at Chinese LLM Cache Pricing

Comments
1 min read
TurboQuant, Four Months Later: Chasing Google's 6x VRAM Claim Into the Wild

TurboQuant, Four Months Later: Chasing Google's 6x VRAM Claim Into the Wild

Comments
6 min read
I built an LLM eval framework from scratch. Here is what I wish I had bought instead.

I built an LLM eval framework from scratch. Here is what I wish I had bought instead.

1
Comments
6 min read
A practical checklist for evaluating an OpenAI-compatible AI API gateway

A practical checklist for evaluating an OpenAI-compatible AI API gateway

Comments
2 min read
I Built an AI Liar's Dice Opponent That Remembers How You Play

I Built an AI Liar's Dice Opponent That Remembers How You Play

5
Comments 5
6 min read
What I learned building a long-lived AI agent (the boring version)

Real-world architecture beats benchmark hype

What I learned building a long-lived AI agent (the boring version)

15
Comments 31
6 min read
2026년 7월 14일 AI·LLM 이슈 다이제스트 — 에이전트가 사무실로 들어오는데, 문단속은 누가 하나

2026년 7월 14일 AI·LLM 이슈 다이제스트 — 에이전트가 사무실로 들어오는데, 문단속은 누가 하나

Comments
1 min read
How I Built a Lightweight Local AI Assistant (Python + PyQt6 + Ollama)

How I Built a Lightweight Local AI Assistant (Python + PyQt6 + Ollama)

Comments
1 min read
Socrates Was the World's First — and Best — Prompt Engineer

Socrates Was the World's First — and Best — Prompt Engineer

1
Comments
8 min read
Nobody audits their OpenAI invoice

Untracked cost gaps reveal leaked keys

Nobody audits their OpenAI invoice

9
Comments 9
5 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.