Home
CV
ExperienceEducation
ProjectsBookmarksInvestmentsContactBlog
X (Twitter)GitHub
X (Twitter)LinkedInGitHubDiscord
Welcome! Type "help" for available commands.
$
Loading terminal interface...
Welcome! Type "help" for available commands.
$
Loading terminal interface...

~/blog/tags/llm

Llm Posts

3 articles

Qwen 3.8 27B Is a Great Open Model

Qwen 3.8 27B Is a Great Open Model

aiqwenllmopen-sourcellama.cpp

A year after OpenAI's gpt-oss, Qwen 3.8 27B is the first local model I'd let write code in an agent harness — fast enough to use live, small enough to self-host, and seemingly twice as capable as last year's open weights on the same hardware.

August 19, 2026
How to Secure Environment Variables for LLMs, MCPs, and AI Tools Using 1Password or Doppler

How to Secure Environment Variables for LLMs, MCPs, and AI Tools Using 1Password or Doppler

security1passworddopplermcpai

Stop hardcoding API keys in MCP configs and AI tool settings. Learn how to use 1Password CLI or Doppler to inject secrets just-in-time for Claude, Cursor, Continue, and other AI coding assistants.

September 25, 2025
Anthropic's Claude 4 Announced: New AI Capabilities for Coding, Agents, and Claude Code in VS Code / IntelliJ

Anthropic's Claude 4 Announced: New AI Capabilities for Coding, Agents, and Claude Code in VS Code / IntelliJ

aiclaudeanthropicllmcodingai-agents

Anthropic launches Claude Opus 4 and Sonnet 4, setting new benchmarks for coding, reasoning, and AI agents with extended thinking capabilities and improved tool use.

May 22, 2025

Discover More