Gemma 4

1. Gemma 4 26B on a consumer GPU: build pain, throughput, and BFCL numbers (algollabs.com via hn)

   +2 8d gemma

   2026-05-05 Gemma 4 26B on consumer-grade 5070Ti GPU A week running Google's Gemma 4 26B as my daily local agent on a single RTX 5070 Ti. No API calls, no cloud, no rate limits.

2. Show HN: I made a Gemma 4 Mac app that names screenshots with local AI (snapname.app via hn)

   +42 9d gemma

   I made my first macOS utility app that ships with a bundled Gemma 4 model, specifically the Gemma E4B one. It made my app DMG have 5.3 GB in size, but I think it is a small size for the power that this free local model can provide.

3. Fun Local LLM Comparisons with Gemma, Granite, and Qwen (ekorbia.com via hn)

   +4 12d gemmaqwen

   Fun local LLM comparisons with Gemma, Granite, and Qwen Ekorbia v0.2 features a comparison-chat mode that runs 2-3 local models against the same prompt in parallel. Here are a few fun prompts running across Gemma 4 (e2b), IBM Granite 4.1 (…

4. Gemma-4-Harmonia-31B-Uncensored-Heretic Is Out Now, a Merge of Multiple gemma-4-31B-it Finetunes Designed for a Targeted Approach to Deep Neural Consolidation, Minimizing Regression While Amplifying Unique Capability Boundaries. With KLD 0.0047 and 9/100 Refusals! (huggingface.co via reddit)

   +134 12d gemma

   Provided in both Safetensors and GGUFs. Safetensors, llmfan46/Gemma-4-Harmonia-31B-it-uncensored-heretic: https://huggingface.co/llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic GGUFs, llmfan46/Gemma-4-Harmonia-31B-it-uncensored-heretic-GG…

5. Running Gemma4 31b-it on vLLM 0.21.0 A100s (bad quality or what am I doing wrong) (www.reddit.com)

   +31 12d vllmgemmaopenai

   Okay fun time I got access to two Nvlinked A100s for some research project I benchmarked my work against the Gemma 4 31b-it available through Google, but my dataset is rather massive, so I need to run it on the "local" resources. Basically…

6. Built a local-first AI memory system that indexes screen activity, meetings, and voice notes ( MCP + automations) (www.reddit.com)

   +17 2w gemmallamacursor+1

   Been experimenting with an idea — what if your AI assistant actually remembered everything you did on your computer? Not stateless chats, but real persistent context.

7. Is something went wrong with those online free model, why I feel they worse than Gemma 4 26B A4B Q4_KM ?? (www.reddit.com)

   +22 2w gemma

   It started with I just want to make a chat app like roleplay with characters but Gemma 4 26B A4B Q4_KM doesn't have info some old character so I crawl back to those online services as those model is much bigger parameter and quite update i…

← all threads