Hi HN, I am Erik Bethke, game dev that has been making online games for 30+ years (Starfleet Command, GoPets, FarmVille, Mafia Wars).

For the last 3 years I have been building Bike4Mind and now I have to put my big boy pants on and 'go to market'. Ugh, it would be so much more cozy to fire up a few more agents and fix a few more bugs, polish some more UX - or hey how about a new feature that does Q, R or S?

What is Bike4Mind? I started Bike4Mind to be the home of a bunch of AI augmented game development tools, and over time have just kept adding features. It is a full cognitive workbench we have all of the features - notebooks, files, data lakes & RAG, agents (different forms), generate share-able artifacts in react or python. We have secure websockets to connect to our CLI - B4M CLI, so it is truly batteries included - and runs any model - OpenAI, Anthropic, Gemini, xAI, Bedrock. Switch models during a conversation, fork a notebook, run 4 different models on the same prompt, loads of image generation support, and import OpenAI and Anthropic history and of course egress with your own at any time.

So, all of the models and all the features. Sure, but so what?

We are huge believers in infrastructure as code and leveraged SSTv4, and Bike4Mind is able to 100% self-host on your own hardware. You can run Qwen locally and have all of the GUI and CLI and be totally offline from the internet.

The most complete fully-sovereign workbench I've found - and if you know a more complete one I'd genuinely like to see it.

Okay, so all of the models, features and I can self host - sounds cool, but why don't I vibe that up myself in Fable or such right?

We have used this with our enterprise clients over the last 3 years and so you also get all of the Enterprise SaaS stuff - batteries included - multi-tenant, RBAC, MFA, throttling, moderation, analytics, premium subscriptions, credits, heck even a slim CRM is built in! OAuth, github, google docs, notion, etc etc… it is one thing to lick the cupcake and get that MVP up and running and working for you - it is a totally different beast putting something that you can deploy for an enterprise.

What? What are you doing exactly?

48 hours ago (July 5th noon Austin) we took everything we built and we made it open with the most progressive license I could lay out: Source immediately under a BSL 1.1 license with a one-way ratchet where after 24 months each release converts to full Apache-2.0. Please appreciate that this is an irrevocable, one-way ratchet where Bike4Mind just gets more open with each release and each month. This is not a "Hey we are open source until well cough reasons - jazz hands we are closed - thanks for all of the PRs!"

To be clear, I know that source-available is not the same as open-source. But check out our actual license and the marketing site - we encourage and permit anyone to fork and create new products and sell them, or take the whole thing and self-host on their own AWS account or their own hardware. You can also fork it and customize a full stack AI solution for your own clients.

Really the only thing we are holding back - and just over 24 months - is not allowing people to stand up a vanilla copy of Bike4Mind and compete with B4M-as-a-service or change the name and do the same.

I built a tiny proxy that gives GLM 5.2 vision (or any text LLM) – MIT

Show HN: Q a REPL for LLM inside the terminal

Code is overpaying for AI. /tokendiet finds where

A New Foundation for Tuxedo OS: Switching to Debian

Sets up your AI agent for Cloudflare

Hudson News self-checkout terminals add 3% "employee benefit" surcharge in NY

Ask HN: Google's chilling effects on expressing democratic dissent

Squish, a local LLM inference server for Apple Silicon

Doing front end changes via LLMs is pain in the a*

Show HN: Xls-reader – read legacy .xls in the browser, no upload, zero deps

Proximity to Golf Courses and Risk of Parkinson Disease

AI Neovim

Show HN: Last EHR – AI agent over a FHIR back end with human approval on writes

Muse Image and Muse Video

Scruter – Ads Intelligence SaaS

Never submit code you don't understand

Ccl: Categorical Configuration Language

Auto: The AGI Compiler

Policy Statement Concerning the Suppression of Accuracy in AI Systems

Tessera - an AI agent that refuses to answer without evidence

Amber the programming language compiled to Bash/Ksh/Zsh

I use AI in creating my newsletter

Muse Image: Free AI Image Generator – Museimage.io

Get Ready for the Powerful CSS Border-Shape Property

Google's exponential path to climate-wrecking digital bloat

Fable Is Good

AI recommends crap travel services

Show HN: Claudeometer – Share and pool Claude usage across a team (macOS, OSS)

Discord banned 8k users for posting benign grid images

Show HN: Bike4Mind – open-core AI workbench; any model, agents, RAG, self-host

Ask HN: What is your AI harness that lets you switch LLM models easily?