LLxprt Code 0.1.16 released - customize and save your model configs, prompts and more. LLxprt Code is a multi-provider, multi-model (including local) LLM coding tool forked from gemini-cli. npm install -g @vybestack/llxprt-code.

A key difference between LLxprt Code and @geminicli besides multi-provider, multi-model, local models, claude style todo lists, better UI handling, customizable prompts, model control and profiles -- is that unlike Gemini CLI -- LLxprt Code is CORGI FREE

Note that with Gemini models, LLxprt Code can already read PDFs (among other files). Moreover, with LLxprt Code you can use far more models from far more providers. Including the mysterious free Horizon Beta on @OpenRouterAI . npm install -g @vybestack/llxprt-code. #justsaying.

Can we talk about how the Windows runners on GitHub are much slower than Linux or Mac despite it being owned by Microsoft?.

LLxprt Code 0.1.15 Released - Better Qwen3-Coder support and Claude Code style TODO lists. LLxprt Code CLI coding tool similar to Claude Code. It is an open source multi-provider multi-model forked from gemini-cli but keeps up with their main as well.

RT @adcock_brett: Here's a F.02 in my home, using Helix to do my laundry

horizon-alpha codes okay. but it is not agentic at all. It wants so many confirmations. I'm struggling to finish the task I set out to do with it. The only reason I believe it is GPT is because GPT loves asking for confirmations too.

Fixed. llxprt --provider openai --keyfile ~/.fireworks_key --baseurl " --model "accounts/fireworks/models/qwen3-coder-480b-a35b-instruct" --prompt "list the files.in this project reponsible for multi-provider communication -- don't look at the docs".

They are deploying a fix. Stay tuned!.

The new Qwen3-Coder model aka Qwen3-Coder-480B-A35B-Instruct is completely broken on @FireworksAI_HQ so don't use it right now. I'd hoped to add it to LLxprt-code, but it emits broken tool calls with unpredictable corruption patterns. @OpenRouterAI works.

LLxprt Code general roadmap. LLxprt Code is a multi-provider/multi-model (including local and open source) fork of Gemini-CLI. It tracks google's releases with additional added features. you can grab 0.1.14 now: .npm install -g @vybestack/llxprt-code

I'm betting I'm in the 5%.

LLxprt is an #opensource multi-provider fork of Google's #gemini-cli. You can still use all the Gemini stuff and tools, but you can also use your own local models as well as other providers like @OpenRouterAI or @FireworksAI_HQ . Yes you can even use @OpenAI or @AnthropicAI.

LLxprt Code 0.1.13 Released. npm install -g @vybestack/llxprt-code@0.1.13. or, upgrade in place.npm update -g @vybestack/llxprt-code. Or use npx.npx @vybestack/llxprt-code. or, run from Docker.docker run -it

I propose the next ECMA spec for JavaScript rename or alias the "any" type to the "claude" type in honor of its biggest fan, user, and proponent.

. but I did find it to be culty.

Something I never knew. Probably because the books at the time weren't updated to mention it. The original XP project failed: The C2 wiki comments are. "cult like" I was heavily influenced by XP early in my career. Still am. .

Since I started working with LLMs I've discovered something really important. Something really the rest of humanity should know about me. I am almost always right! . Claude, o3, GPT all agree -- one among humans is almost always correct and it is me!.

Officially launched: LLxprt-Code. Use one CLI to use any model or provider that you like. LLxprt-code is a community fork of gemini-cli.npm install -g @vybestack/llxprt-code.

RT @Corix_JC: It’s time to completely change how #data management works . . @InfoWorld @ACOliver #tech #digital #in….