Hackernews posts about Gemma

1. Fable 5 pushed Gemma 4 to 255 tok/s on WebGPU (xcancel.com)
   48 points by kirubakaran 18 days ago | 22 comments
2. Gemma 4 on Cerebras - The Fastest Inference Is Now Multimodal (www.cerebras.ai)
   24 points by Tiberium 7 days ago | 8 comments
3. Gemma 4 E2B running in-browser at 255 tok/s (huggingface.co)
   19 points by victormustar 19 days ago | discuss
4. Show HN: Local privacy-first Microsoft Recall alternative with Gemma 4 (github.com)
   13 points by skye0110 3 days ago | 2 comments
5. Show HN: Running Gemma-4 26B at 124 tokens/SEC on a CPU, no GPU (apeg.dev)
   10 points by arun-prasath 6 days ago | 1 comments
6. Show HN: Gemma 3 inference in pure C++ with Metal acceleration (github.com)
   6 points by ybubnov 2 days ago | 2 comments
7. Faster Gemma 4 on MLX with multi-token prediction (ollama.com)
   6 points by ilreb 6 days ago | discuss
8. Gemma 4 in the browser with WebGPU (huggingface.co)
   5 points by astlouis44 4 days ago | discuss
9. Gemma 4 for Telephony: From Two AI Models to One – Until I Switched to Chinese (medium.com)
   4 points by fidotron 22 days ago | discuss
10. When Gemma Thinks About Resources – It Fails: A Behavioral Experiment (www.lesswrong.com)
    3 points by joozio about 13 hours ago | discuss
11. Local LLMs on a Ryzen 8700G iGPU: 13-15 tok/s on gemma4, 9-12 on qwen3.6 (loomcycle.dev)
    3 points by denn-gubsky 6 days ago | discuss
12. I tested Gemma4 12B on my 8GB GPU, now I don't want to go back to smaller models (www.xda-developers.com)
    3 points by theanonymousone 18 days ago | discuss
13. Bonsai: Human->LLM->Web with LLM interface using Gemma4 12B locally on Windows (drive.google.com)
    2 points by coolwulf 27 days ago | 1 comments
14. Show HN: GemmaTrans – On-device translation for macOS, built with MLX (github.com)
    2 points by Rand01ph 9 days ago | discuss
15. Run Gemma on the edge with the Coral Board [video] (www.youtube.com)
    2 points by simonpure 20 days ago | discuss
16. Gemma-4-31B at 256K context on a $1,400 AMD GPU – measured, with patches (github.com)
    2 points by alsaha 26 days ago | discuss
17. Gemma 4 E4B as a primary local LLM (replaced Qwen) (digg.com)
    2 points by galsapir 29 days ago | discuss
18. Show HN: peerd – AI agent harness that runs entirely in your browser (github.com)
    75 points by NotASithLord 13 days ago | 23 comments
19. Show HN: Run Agent Skills with mistral.rs v0.8.10: /v1/skills support and more
    20 points by ericlbuehler 19 days ago | 2 comments
20. Ask HN: Any Local LLM can I run without GPU for Local Agentic workflow AI?
    8 points by limondas 26 days ago | 4 comments
21. Ask HN: How is GPU power draw measured at scale?
    6 points by anax32 10 days ago | 2 comments
22. Ask HN: Is it feasible to run a model on device for complete privacy?
    3 points by mazinz 29 days ago | 6 comments
23. Show HN: Loqi, a "local-first" translation tool using Ollama/llama.cpp (github.com)
    3 points by danterolle 14 days ago | discuss
24. Show HN: CriteriaBot – A Universal Customizable Classifier (criteriabot.io)
    3 points by RoyalTnetennba 21 days ago | discuss
25. Show HN: Private Wealth Tracker (apps.apple.com)
    2 points by mazinz 26 days ago | discuss
26. Show HN: Arda Translate, free private on-device (iOS) translation (apps.apple.com)
    1 points by cjjuice about 2 hours ago | 1 comments
27. Ask HN: What are you go to LLM models for the following
    1 points by mrburton 4 days ago | 1 comments
28. Show HN: Cerver is infra for AI sessions (cerver.ai)
    1 points by eyalgoren 13 days ago | discuss
29. Show HN: Language1 – Benchmarking LLM comprehension of vague prompts via Taboo (language1.app)
    1 points by kaandemirel 18 days ago | discuss
30. DiffusionGemma: 4x Faster Text Generation (blog.google)
    327 points by meetpateltech 26 days ago | 88 comments