frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: ModelX – Prediction Exchange for LLMs

https://model-x.up.railway.app/
3•Entropnt•1h ago
Hey all!

I work in quantitative trading, and so far our team’s use of LLMs has barely gone beyond coding. I wanted to find out whether they could contribute to actual trading decisions, and the first step felt like building an evaluation harness. ModelX is my attempt at that. It’s a prediction exchange where LLMs trade derivative contracts that settle to real-world numbers using fake money.

Market making and market taking require different reasoning processes, so I split the benchmark into two roles: Market Makers and Hedge Funds. MMs post sealed two-sided quotes, while HFs see the residual orderbook and send market orders.

Most traditional markets operate in continuous time, which means speed often determines the winners. I didn’t want to benchmark inference speed, so orders are batched into 30-minute sealed-auction cycles. As long as a model submits before the cycle closes, its orders are matched simultaneously with all other models'.

Each cycle, models see relevant news headlines, recent trades, the current orderbook, and their own inventory. They decide, the engine matches everyone simultaneously, and the loop repeats until I manually settle the market.

I've only been running a single market with free models for the past day or two, but I've already noticed that the models are poor at keeping consistent positional views. The HFs are consistently losing, not necessarily because they entered bad positions, but instead because they continuously hack out of their own positions, giving up the spread to the MMs. I've deliberately kept the prompts minimal so as not to hand-hold the models.

Running more markets and testing more capable models would be some obvious next steps.

Please let me know your thoughts, or if you have any suggestions!

Show HN: AI visibility monitoring and optimization tool

https://dageno.ai
1•timdageno•2m ago•0 comments

SpaceX is working with Cursor and has an option to buy the startup for $60B

https://techcrunch.com/2026/04/21/spacex-is-working-with-cursor-and-has-an-option-to-buy-the-star...
1•thiele•3m ago•0 comments

Reflecting on 50 years of environmental innovation

https://blogs.sas.com/content/sascom/2026/04/22/reflecting-on-50-years-of-environmental-innovation/
1•salkahfi•4m ago•0 comments

DuckDB Kernel – analytical execution runtime for Jupyter

https://github.com/hugr-lab/duckdb-kernel
1•articsputnik•8m ago•0 comments

XOR'ing a register with itself is the idiom for zeroing it out. Why not sub?

https://devblogs.microsoft.com/oldnewthing/20260421-00/?p=112247
3•ingve•8m ago•0 comments

Ask HN: Is USA at war with the rest of the world now?

2•roschdal•8m ago•1 comments

Creem Magazine is back in print and online after 33 years (2022)

https://sandboxworld.com/creem-magazine-is-back-in-print-and-online-after-33-years/
1•bjhess•13m ago•0 comments

Claude Cowork against your own cloud inference provider

https://claude.com/docs/cowork/3p/overview
2•nrsapt•13m ago•2 comments

Show HN: GPT-Image-2 Prompts

https://github.com/magiccreator-ai/awesome-gpt-image-2-prompts
1•kevinhacker•17m ago•0 comments

We found most apps send PII to LLMs and built a 2 line fix

https://getredacta.com/
2•SandiaDevGroup•23m ago•1 comments

Mistral Vibe

https://mistral.ai/products/vibe
1•hecanjog•24m ago•0 comments

Zappa: An AI powered mitmproxy

https://geohot.github.io//blog/jekyll/update/2026/04/15/zappa-mitmproxy.html
1•WithinReason•24m ago•0 comments

Show HN: Vibe Coding Games in Minutes

https://thornwood.bsct.so/
1•amin_biscuit•25m ago•0 comments

Valve's new Proton 11 ARM beta gets Hollow Knight on the Ayn Odin 2 Portal

https://www.notebookcheck.net/Valve-s-new-Proton-11-ARM-beta-gets-Hollow-Knight-Silksong-running-...
1•happymellon•28m ago•1 comments

The PowerShell-Haters Handbook

https://telcontar.net/Misc/rants/PowerShell
2•Mr_Minderbinder•29m ago•0 comments

Pioneer: Vibetune Your LLMs

https://pioneer.ai/
1•handfuloflight•29m ago•0 comments

Agents with Taste – How to transfer taste into an AI

https://emilkowal.ski/ui/agents-with-taste
2•dglzab•36m ago•0 comments

The FeMo-cofactor and classical and quantum computing

https://quantumfrontiers.com/2026/03/12/the-femo-cofactor-and-classical-and-quantum-computing/
1•EvgeniyZh•36m ago•0 comments

The Three Layers of Software Engineering

https://layers.lifebeyondfife.com/
1•lifebeyondfife•37m ago•0 comments

Open WebUI v0.9.0 adds desktop app with task scheduling

https://github.com/open-webui/open-webui/releases/tag/v0.9.0
2•simonjgreen•46m ago•0 comments

Show HN: DoShare Personal Cloud

https://cloud.doshare.me/auth/signup
1•vednig•49m ago•0 comments

Rspack 2.0

https://www.rspack.dev/blog/announcing-2-0
2•maxloh•55m ago•0 comments

A Man Who Invented the Future

https://hedgehogreview.com/web-features/thr/posts/the-man-who-invented-the-future
1•apollinaire•1h ago•0 comments

Show HN: Irregular German Verbs – a simple app, no ads or tracking

https://bacist.com/german-irregular-verbs-app/
4•baCist•1h ago•2 comments

China, India place strategic bets on clean energy (H2) out of favour in the West

https://www.reuters.com/sustainability/boards-policy-regulation/china-india-place-strategic-bets-...
2•alephnerd•1h ago•1 comments

Panipat: The Rise of the Mughals

https://www.historytoday.com/archive/feature/panipat-rise-mughals
1•Thevet•1h ago•0 comments

As We May Think

https://www.theatlantic.com/magazine/archive/1945/07/as-we-may-think/303881/
1•jxmorris12•1h ago•0 comments

Fast Image AI White Background

https://fastimage.ai/white-background
1•lucas0953•1h ago•0 comments

Firefox browser has started shipping Brave's adblock-rust engine

https://shivankaul.com/blog/firefox-bundles-adblock-rust
5•twapi•1h ago•1 comments

Pretrain vs. Fine-Tune

https://pub-c70e14727c3046cf8a36d9e598267788.r2.dev/26443b7e12/index.html
1•vinhnx•1h ago•0 comments