DEV Community

#llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
How to Run LLMs Locally When Cloud AI Gets Too Invasive

How to Run LLMs Locally When Cloud AI Gets Too Invasive

Comments
5 min read
Most document AI questions aren't retrieval problems

Most document AI questions aren't retrieval problems

4
Comments
4 min read
How I got 80% code retrieval accuracy without vectors, embeddings, or any ML

How I got 80% code retrieval accuracy without vectors, embeddings, or any ML

Comments
2 min read
AI-generated accessibility, an update — frontier models still fail, but skills change the game

AI-generated accessibility, an update — frontier models still fail, but skills change the game

4
Comments 3
6 min read
Agentic AI's Infrastructure Boom Meets Its Reliability Problem

Agentic AI's Infrastructure Boom Meets Its Reliability Problem

Comments
3 min read
Why I built ragwise: pip-installable RAG with hybrid search, streaming, and agent tools by default

Why I built ragwise: pip-installable RAG with hybrid search, streaming, and agent tools by default

Comments
4 min read
The Real Problems Start After Your MCP Server Works

The Real Problems Start After Your MCP Server Works

4
Comments 9
3 min read
What "Subquadratic Attention" Actually Means

What "Subquadratic Attention" Actually Means

Comments
4 min read
Request-Boundary AI Spend Control in 2026: A Practical Diagnostic for Gateway and FinOps Teams

Request-Boundary AI Spend Control in 2026: A Practical Diagnostic for Gateway and FinOps Teams

1
Comments 6
5 min read
All Data and AI Weekly #238-20April2026

All Data and AI Weekly #238-20April2026

5
Comments
11 min read
When one translation isn't enough: building konid

When one translation isn't enough: building konid

Comments
2 min read
I was embarrassed by my RAG demo. Turns out the bug was never in my code.

I was embarrassed by my RAG demo. Turns out the bug was never in my code.

Comments 1
3 min read
Opus 4.7 First Look: I Tested the Day-Old Model Against 3 Other Claudes on 10 Real Tasks

Opus 4.7 First Look: I Tested the Day-Old Model Against 3 Other Claudes on 10 Real Tasks

Comments 1
5 min read
I Built a 7-Agent Prompt Framework, Then Used It to Debug Its Own Output

I Built a 7-Agent Prompt Framework, Then Used It to Debug Its Own Output

Comments
6 min read
LLM routing per tier via OpenRouter — when one model doesn't fit all

LLM routing per tier via OpenRouter — when one model doesn't fit all

Comments
6 min read
đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.