redwood_

Agree, these new models are a game changer. I switched from Claude to Qwen3-Coder-Next for day-to-day on dev projects and don't see a big difference. Just use Claude when I need comprehensive planning or review. Running Qwen3-Coder-Next-Q8 with 256K context.

I run Qwen3-Coder-Next (Qwen3-Coder-Next-UD-Q4_K_XL) on the Framework ITX board (Max+ 395 - 128GB) custom build. Avg. eval at 200-300 t/s and output at 35-40 t/s running with llama.cpp using rocm. Prefer Claude Code for cli.

Yes just local, but love to use Nao to quickly analyze datasets

Like the looks of this. Any chance you'll be adding support for SQLite at some point?

Stealth Startup | Boston (US) | Onsite/Hybrid | Full-Time

Stealth is a fintech startup with a vision to speed up liquidity in the private markets. Founding team has prior exit in the space and has secured funding to do it again. We leverage domain expertise and network to tackle the lack of data and high transaction costs in the market. Technology is a combination of cutting-edge AI/ML, async processing and organizing of unstructured datasets and workflow tooling.

You'll be employee #3 with a chance to go fast and influence product, code and culture.

We're hiring for one Full-Stack Engineer. Experience with React/Angular, Python, Postgres/Mongo, DevOps (RDS, Lambda) is an asset.

Please email us (dev_position at sekond dot co) and mention HN in the subject.

Yes, they just did start buying from Russia. It is a security concern these days and was discussed as recently as yesterday in Swedish media.