DEV Community

#llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
Mellum2 MoE, Heretic Censorship Removal, & NVIDIA Cosmos 3 Omni-model for Local AI

Mellum2 MoE, Heretic Censorship Removal, & NVIDIA Cosmos 3 Omni-model for Local AI

Comments
3 min read
Claude Code detected a hack that never happened, then spiraled

Claude Code detected a hack that never happened, then spiraled

4
Comments 5
5 min read
A weekly tool-use test is the only signal

A weekly tool-use test is the only signal

Comments
3 min read
I built an llms.txt for Salesforce — so AI stops writing deprecated Apex

I built an llms.txt for Salesforce — so AI stops writing deprecated Apex

Comments
4 min read
How Meta's AI Support Bot Got Tricked Into Hijacking Instagram Accounts

How Meta's AI Support Bot Got Tricked Into Hijacking Instagram Accounts

1
Comments
5 min read
Provider drift broke our regression evals. We pinned versions through Bifrost.

Provider drift broke our regression evals. We pinned versions through Bifrost.

Comments
4 min read
What every LLM call in your Flutter app actually costs

What every LLM call in your Flutter app actually costs

Comments
1 min read
Centralising tool access for our prompt-assembly agent with Bifrost MCP gateway

Centralising tool access for our prompt-assembly agent with Bifrost MCP gateway

Comments
4 min read
How We Cut AI Infrastructure Costs by 94% Without Sacrificing Quality (And How You Can Too)

How We Cut AI Infrastructure Costs by 94% Without Sacrificing Quality (And How You Can Too)

Comments
13 min read
What I Learned Building a Local RAG Agent

What I Learned Building a Local RAG Agent

Comments
5 min read
Cache-hit dispersion is the 7th vendor-risk axis — and the one your invoice can't see

Cache-hit dispersion is the 7th vendor-risk axis — and the one your invoice can't see

Comments
8 min read
I let GPT-4o and a cheaper model fight over my inbox. GPT-4o lost.

I let GPT-4o and a cheaper model fight over my inbox. GPT-4o lost.

8
Comments 5
4 min read
One image schema for four VLM providers: we stopped reformatting payloads

One image schema for four VLM providers: we stopped reformatting payloads

Comments
4 min read
#3 setting env

#3 setting env

Comments
2 min read
Rails, Not Rules: enforcing a coding agent's domain vocabulary with checks

Rails, Not Rules: enforcing a coding agent's domain vocabulary with checks

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