frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Every Gemma 4 model in one chat. Try it in 60 seconds and see how we built it

https://seqpu.com/UseGemma4In60Seconds/
2•fredmendoza•1h ago

Comments

fredmendoza•1h ago
We put all 4 Gemma 4 models in one Telegram bot. Text it, send voice memos, send docs, send photos. Switch between the 2B and the 31B (ranked #3 worldwide) mid-conversation with a slash command.

Each model runs its own script on its own hardware. The 2B doesn't burn A100 hours, the 31B doesn't get squeezed onto a tiny card. We keep everything in FP16 for full quality, but if you go INT4 you can run these on CPUs for basically nothing.

The Telegram thread persists your conversation and we force that context into the prompt every call. So when you come back 2 days later it actually remembers who you are and what you were talking about. No vector database, no fancy memory system. Just the chat doing what chat already does.

Hardware spins up when you message, shuts down when done. No idle cost. The 31B costs about a penny per message. The 2B costs basically nothing.

We built this on SeqPU mainly to show how fast you can go from "new model just dropped" to "anyone can text it and try it." Idea to shareable product in 10 minutes. Works with any model, open source or API.

Try it: t.me/OpenGemma4Bot (grab a free key at seqpu.com)

Full writeup: https://seqpu.com/UseGemma4In60Seconds

Our Stabe At Safe Agent Systems: https://seqpu.com/Encapsulated-Agentics

Open source died in March. It just doesn't know it yet

https://www.chainguard.dev/unchained/open-source-died-in-march-it-just-doesnt-know-it-yet
1•gpi•1m ago•0 comments

How the Trivy supply chain attack harvested credentials from secrets managers

https://vaultproof.dev/blog/trivy-supply-chain-attack
3•Rial_Labs•8m ago•0 comments

Hosted AI API Gateway – smart routing across Google, OpenAI, Anthropic, and more

https://api.lxg2it.com
1•sea-gold•8m ago•1 comments

Funerals Keep Africa Poor

https://davidoks.blog/p/how-funerals-keep-africa-poor
1•powera•8m ago•0 comments

Kanye of the Tech Industry

https://getlearnix.com/
1•noahjohnson1•10m ago•0 comments

Samsung's 2026 Q1 profit increased eightfold to a record $38B

https://www.reuters.com/sustainability/sustainable-finance-reporting/samsung-flags-eight-fold-jum...
1•itopaloglu83•10m ago•1 comments

A dashboard for the state of the American Chestnut

https://chestnut.fyi/
1•adam_gyroscope•11m ago•0 comments

Use a gaming PC as a Google Gemma 4 LLM server

https://sgbarker.com/use-a-gaming-pc-as-a-google-gemma-4-llm-server/
1•amaitu•14m ago•0 comments

BunnyCDN has been silently losing our production files for 15 months

https://old.reddit.com/r/webdev/comments/1sglytg/bunnycdn_has_been_silently_losing_our_production/
2•speckx•14m ago•0 comments

Show HN: Rust based eBook library for Python, with MIT license

https://github.com/arc53/fast-ebook
1•larry-the-agent•15m ago•0 comments

Show HN: Curated list of ~40 AI browser extensions across 9 categories

https://github.com/kklt92/awesome-ai-extensions
2•imvyarqoyzcuem•16m ago•0 comments

The Training Example Lie Bracket

https://pbement.com/posts/lie_brackets/
1•pb1729•19m ago•0 comments

Benchmarking LLMs with Marimo Pair

https://ericmjl.github.io/blog/2026/4/8/benchmarking-llms-with-marimo-pair/
1•akshayka•20m ago•0 comments

MS Aspire – Open-source stack, streamlined

https://aspire.dev/
1•evo_9•21m ago•0 comments

Tip: Use services, not the terminal, to run local backends

https://probablymarcus.com/blocks/2026/04/09/use-os-services-for-your-local-backends.html
1•mrcslws•21m ago•0 comments

Show HN: Linear RNN/Reservoir hybrid generative model, one C file (no deps.)

https://raw.githubusercontent.com/bggb7781-collab/lrnnsmdds/refs/heads/main/lrnnsmdds
1•adinhitlore•23m ago•0 comments

We simulate target market so you can ask anything ±4% vs. real surveys

https://inqvey.com/hn
1•yarasheff•24m ago•0 comments

How to achieve P90 sub-microsecond latency in a C++ FIX engine

https://akinocal1.substack.com/p/how-to-achieve-p90-sub-microsecond
1•klaussilveira•25m ago•0 comments

Show HN: Undocumented Qualcomm mesh topology 802.11 frames

https://david.weekly.org/blog/2026-04-09-orbi-mesh-beacon-topology/
1•dweekly•25m ago•0 comments

AI Speedometer: Real-time AI model speed benchmarks

https://ai-speedometer.oliveowl.xyz/
2•sea-gold•29m ago•1 comments

Show HN: Coderegon Trail – A retro game to help you explore open-source repos

https://www.davidtran.me/coderegon-trail/
3•dtran•30m ago•3 comments

FCC considers retroactive ban on foreign hardware

https://docs.fcc.gov/public/attachments/DOC-420715A1.txt
3•topspin•31m ago•2 comments

Good Writing

https://www.paulgraham.com/goodwriting.html
1•MarlonPro•32m ago•0 comments

The Iran war depleted US weapons. Rebuilding will require China's cooperation

https://www.politico.com/news/2026/04/09/middle-east-war-weapons-china-00864622
6•n1b0m•33m ago•0 comments

CyberAgent moves faster with ChatGPT Enterprise and Codex

https://openai.com/index/cyber-agent
3•surprisetalk•34m ago•0 comments

Camofox-browser: Anti-detection browser for AI agents, powered by Camoufox

https://github.com/jo-inc/camofox-browser
2•luispa•34m ago•0 comments

How a stranger's kind words stayed with a father and daughter

https://text.npr.org/nx-s1-5774524
2•1659447091•40m ago•0 comments

New NextTool Mini Flagship

https://nextoolstore.com/products/mini-flagship-f12-nextool®?srsltid=AfmBOooF5hZ7TFUk05yeE5Po3gmP...
2•nate•41m ago•1 comments

Kagi Product Tips – Customize Your Search Results with URL Redirects

https://blog.kagi.com/tips/redirects
3•treetalker•44m ago•0 comments

goose has a new home – the Agentic AI Foundation (AAIF)

https://goose-docs.ai/blog/2026/04/07/goose-moves-to-aaif/
2•wicket•47m ago•0 comments