Hey, We started Offline Base because: 1. We believe Local[ly runnable], Open Source, and Small Language Models are good enough to do just about everything that people use frontier models for. 2. We believe that running models locally is the solution to the many problems associated with or attributed to data centers and cloud based models 3. We believe that teaching everyone about data privacy is a lost cause (sort of) 4. We want to ensure that no one gets left behind in this new AI-Native World 5. We believe that Open Source models are more than good enough. We hope to make Local AI easy for all.

love to see it 👏

Had a good time going to the Open by Design event hosted by @reflection_ai @nvidiaai @togethercompute with @electricaliee If you’re interested in local, edge, open source ai, let’s chat!

@MiniMax_AI 3.1 available on tryingopen now

Working on using @liquidai d1 to help you moderate your, your kids, your employees, etc. before things get sent to AI models, infra, etc. All done locally! Working on this and more @offlinebase

damn! actual “decision tree” with Liquid d1 🌱🌱🌱🌱🌱🌱🌱🌱🌱🌱🌱🌱🌱🌱🌱

Actual “decision tree”. @Raspberry_Pi 5 running @liquidai d1 600m Being powered by solar With a diy tree connected to Figuring out how to productize this for < $125

Insane work by @reflection_ai

Try deepseek v4.1 Flash 1017 here on tryingopen[.]com for free @deepseek_ai

Today we’re releasing Muse Spark Open, the most capable US open-weight model to date. Built for coding and agentic work, it excels at long-horizon tasks, complex software engineering, and multi-step reasoning. Key Model Specs: → 816B total / 42B active parameters → 1M token context window → Natively multimodal across text, image, video, audio

Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about Beam: reflection.ai/beam

🚀 DeepSeek-V4.1-Flash-1017 is here! Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API. 🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model. 1/4

Try out @liquidai new LFM 2.5 24B model routed by their D1 decision model, made by us @offlinebase. Website in comments. @EthanCastr

If anyone wants to try @Muse Spark Open for free with no logins tryingopen[.]com

This is fire

We are taking down tryingcoreweave.com by request of the @CoreWeave legal team. It was a fun run and I am glad many of you got to try their amazing infra through the site.

If you want to try @NVIDIAAI Nemotron 3.5 Super VL for free, check out tryingopen[.]com

Embedding Gemma 2 routing you between Gemini 4 Argon and Gemma 4 e2b. Check it out. Thanks for the insane release(s) @OfficialLoganK @Google @GoogleAI @sundarpichai @GlennCameronjr @googlegemma @GeminiApp @GoogleAIStudio @Electricaliee @GoogleLabs (let's make this an official experiment!) Check out GemmaGemini[.]com

Moving tryingopen.com model router to d1 by @liquidai running on jetson orin nanos @NVIDIAAI

what is tryingopen piped.video/aCVI5bbbEgs?si=ZbIE…

Self host AI chats on your device or your google drive. One click, no login, zero data retention, free. Try it on Trying Open from @offlinebase, link in thread We're trying to figure out ways to improve privacy and reduce the amt of places your data lives.