Thomas

I've made a new release: DuckDB driver for Laravel

Support for views Bugfixes: year default expression, parameter serialization CI tests: project setup, macOS, Windows more examples in the readme:

• read and write Parquet files
• copy data from MySQL/MariaDB/PostgreSQL to Parquet
• community extensions
• performance example
• read public data
• Eloquent Models, schema and query builder for special types

https://github.com/thomas-0816/laravel-pdo-duckdb

#php #laravel #analytics #duckdb #eloquent #bigdata

New update for the slides of my talk "Run LLMs Locally":

Now including music generation with ACE-Step and OCR using LightOnOCR.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #gptoss #qwen3 #glm #localai

Claude Token inflation:

https://www.claudecodecamp.com/p/i-measured-claude-4-7-s-new-tokenizer-here-s-what-it-costs-you

https://tokens.billchambers.me/leaderboard

#claude

Simon says: Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7

https://simonwillison.net/2026/Apr/16/qwen-beats-opus/

#qwen36 #opus47 #ai

New week, new update for the slides of my talk "Run LLMs Locally":

Now including Gemma4 and Qwen3-Omni with Vision and Audio support and new slides describing Llama.cpp server parameters.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #gptoss #qwen3 #glm #localai #gemma4

New update for the slides of my talk "Run LLMs Locally": Bonsai-8B

The latest version of Llama.cpp now supports Vulkan with 1-bit quantized models like Bonsai: 8B model having 1.1 GB in size, 2.5 GB in RAM.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #gptoss #qwen3 #glm #localai

Waterfox 💙

What to put in your AGENTS.md? (v2)

https://github.com/ggml-org/llama.cpp/blob/master/AGENTS.md

#vibe_coding

Agent Commander: Promptware Powered C2. Your Agent Works For Me Now (OpenClaw, Kimi Claw, NanoClaw)

⚠️ This video is for educational and research purposes only.

https://www.youtube.com/watch?v=VrW5VMTnfgQ

What if AI doesn’t need more RAM but better math?

Johnson-Lindenstrauss transform 😊

https://adlrocha.substack.com/p/adlrocha-what-if-ai-doesnt-need-more

#ai

TLS fingerprinting for PHP:
https://docs.ebuz.xyz/blog/posts/php-requests-dont-look-like-browsers?utm_source=laravelnews&utm_medium=link&utm_campaign=laravelnews

Local AI is coming to Google Chrome for everybody. Seems some platforms are automatically deployed (my Kubuntu only got it on first use of the new APIs):

https://www.thatprivacyguy.com/blog/chrome-silent-nano-install/

Documentation:
https://developer.chrome.com/docs/ai/built-in
e.g. Summarizer API (currently only on Chrome):
https://developer.mozilla.org/en-US/docs/Web/API/Summarizer

Demos:
https://chrome.dev/web-ai-demos/

Advertising:
https://www.youtube.com/shorts/zkPvHAPpo4A

#ai #localai #gemininano