frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: ModelAtlas – Find AI models that HuggingFace search can't

https://github.com/rohanvinaik/ModelAtlas
1•Vinaik•1h ago

Comments

Vinaik•1h ago
I built this because I kept having the same experience: I knew the kind of model I wanted, but HuggingFace couldn't express my query.

You search for "small code model with tool-calling." HF gives you Qwen2.5-Coder-32B and a 480B model. Sorted by popularity. It can't express "small" as a direction — it can only filter by tags and sort by likes.

ModelAtlas treats model space as a coordinate system. Eight signed dimensions (architecture, capability, efficiency, domain, etc.) plus 166 semantic anchors. You query by direction: "I want something small (+efficiency), code-focused (+capability), with tool-calling." The scoring is multiplicative — a model that nails efficiency but misses capability gets zero, not fifty percent.

The README has a three-level comparison against HuggingFace search (same queries, both systems, real results):

--Level 1: Common queries — both systems return the right models. This is the baseline.

--Level 2: Directional queries ("small," "fast," "medical classifier") — HF starts returning noise. "Fast embedding model" returns nothing because "fast" isn't a tag.

--Level 3: Multi-constraint queries ("multilingual chat, NOT code/math") — HF can't express these at all. ModelAtlas finds a 100M-parameter TTS model, a genomics classifier, a 0.8B model distilled from Claude Opus. All real models on HuggingFace, all invisible to keyword search.

It's an MCP tool — works inside Claude Code, Cursor, VS Code, any MCP client. The idea is to give a model that calls the tool a kind of subconscious vibe about what's out there on HuggingFace. Instead of guessing from stale training data or making you browse HF yourself, the LLM gets structured awareness of ~30K models in one tool call — ~500 tokens, <100ms. It just knows what model fits your use case.

The network ships as a pre-built SQLite file via GitHub Releases. It covers the community-validated portion of HuggingFace (models with 10+ likes) and is mostly up to date — this is a fun personal project so it's not always updated to the minute, but it's MIT licensed so anyone can run the ingestion pipeline themselves or fork it however they want.

Repo: https://github.com/rohanvinaik/ModelAtlas

Happy to answer questions about the scoring math, the extraction pipeline, or anything else!

I published the architecture Anthropic is now shippingstamped proof

https://nicholasjbogaert.substack.com/p/i-published-the-architecture-anthropic
2•nicbogaert•1m ago•0 comments

TMZ Has a New Obsession: Vacationing Members of Congress

https://www.notus.org/congress/tmz-congress-coverage-dhs-shutdown-tabloid-paparazzi
2•OrangePilled•2m ago•0 comments

Functional hierarchy of the human neocortex across the lifespan

https://www.nature.com/articles/s41586-026-10219-x
1•bookofjoe•2m ago•0 comments

Redesign any website with four AI models side by side

https://shuffle.dev/ai-website-redesign
2•avoc777•3m ago•0 comments

Cisco source code stolen in Trivy-linked dev environment breach

https://www.bleepingcomputer.com/news/security/cisco-source-code-stolen-in-trivy-linked-dev-envir...
4•_____k•6m ago•0 comments

Scaling Laws: The Lawfare Playlist on AI [video]

https://www.youtube.com/playlist?list=PL9f-8IUHQF3lNz1hhIIU1AyQM9M6hmP4m
1•verdverm•6m ago•0 comments

Ministack (Replacement for LocalStack)

https://ministack.org/
3•kerblang•8m ago•0 comments

APS: Open specification for AI agent policies

https://agentpolicyspecification.github.io
1•pascalwilbrink•10m ago•0 comments

Show HN: AI writes the code, but who answers in the interview?

https://sharpskill.dev/en
1•GiornoJojo•10m ago•0 comments

Amazon pulled Kindle update 5.19.3 a few days after release

https://old.reddit.com/r/kindle/comments/1s7lo50/5193_officially_back_to_5192_on_kindle_website/
2•seam_carver•11m ago•0 comments

OpenAI Valued at $852B After Completing $122B Round

https://www.bloomberg.com/news/articles/2026-03-31/openai-valued-at-852-billion-after-completing-...
2•matthieu_bl•11m ago•0 comments

Top Five (Wrong) Reasons You Don't Have Testers (2000)

https://www.joelonsoftware.com/2000/04/30/top-five-wrong-reasons-you-dont-have-testers/
3•Austin_Conlon•12m ago•0 comments

Codex Plugin for Claude Code

https://community.openai.com/t/introducing-codex-plugin-for-claude-code/1378186
1•grodriguez100•14m ago•0 comments

Show HN: Fixing Claude Code's amnesia with persistent memory

https://github.com/agynio/claude-map-reduce-memory
1•NBenkovich•17m ago•2 comments

OpenHarness – Composable TypeScript SDK for building powerful agent harnesses

https://www.open-harness.dev
2•MaxGfeller•17m ago•1 comments

How I Fixed My Headaches

https://chillphysicsenjoyer.substack.com/p/how-i-fixed-my-headaches
4•sebg•18m ago•0 comments

CLaudeCode Feed Keywords

https://github.com/chatgptprojects/claude-code/blob/642c7f944bbe5f7e57c05d756ab7fa7c9c5035cc/src/...
1•oriettaxx•18m ago•0 comments

Open Swarm, open source platform for running AI agents in parallel

https://github.com/openswarm-ai/openswarm
2•ciregenz10•22m ago•1 comments

A personal knowledge graph builder, applied to Bible study

https://www.bibletrace.com
3•vandervs•23m ago•1 comments

Which Countries Depend the Most on Persian Gulf Oil and Gas

https://www.nytimes.com/interactive/2026/03/23/business/energy-environment/oil-importers-middle-e...
2•speckx•23m ago•0 comments

Pidgin 3.0 Alpha 1 (2.95.0) has been released

https://discourse.imfreedom.org/t/pidgin-3-0-alpha-1-2-95-0-has-been-released/378
2•rw_grim•26m ago•0 comments

Sanity-checking ML metrics before trusting them

https://predictly.cloud
1•johndreynolds•27m ago•0 comments

America Got a Little More Religious. No, Really.

https://www.graphsaboutreligion.com/p/america-got-a-little-more-religious
3•sebg•28m ago•1 comments

Formal Verification in Any Language for Everybody

https://www.dev-log.me/formal_verification_in_any_language_for_everybody/
4•yannick-cw•28m ago•0 comments

Safe ways to do things in bash (2023)

https://github.com/anordal/shellharden/blob/master/how_to_do_things_safely_in_bash.md
3•gautamsomani•28m ago•0 comments

Every Package You Install Can Read Your Secrets

https://www.eliranturgeman.com/2026/03/28/supply-chain-attacks/
2•birdculture•28m ago•0 comments

ClaudeDown: Is Claude getting dumber, or is it just you?

https://claudedown.com
3•prabal97•30m ago•1 comments

Mercury Edit 2: Fastest next-edit prediction with a diffusion LLM (221ms)

https://www.inceptionlabs.ai/blog/introducing-mercury-edit-2
1•nathan-barry•33m ago•0 comments

Show HN: I vibe coded Zeitbook – create custom books from any idea

https://zeitbook.com/
2•makossa•34m ago•0 comments

Backstage Is Dead

https://autonomousengineering.substack.com/p/backstage-is-dead
5•krakenwake•34m ago•0 comments