VQV Signal
According to a post on X by @thinkymachines, Inkling 2 Small's weights are now available under the Apache 2.0 license. The model is a 594B-parameter mixture-of-experts system with 17B parameters active and up to 1M context. It is on Hugging Face and live on Together, Modal, Ollama and others. The post says Arena ranks it as the #1 US open model in Code Arena WebDev.
According to a post on X by @thinkymachines, Inkling 2 Small's weights are now available under the Apache 2.0 license. The model is a 594B-parameter mixture-of-experts system with 17B parameters active and up to 1M context. It is on Hugging Face and live on Together, Modal, Ollama and others. The post says Arena ranks it as the #1 US open model in Code Arena WebDev.
AI-assisted summary based on the listed source.
The 594B-parameter mixture-of-experts model (17B active, up to 1M context) is now on Hugging Face and live on Together, Modal, Ollama and others, and Arena ranks it the #1 US open model in Code Arena WebDev. It gives builders a top-ranked, commercially usable US open-weight option, though it needs roughly 500GB of...
Builders looking for a US-made open-weight model they can use commercially now have another option to evaluate. Running it needs roughly 500GB of GPU memory, which limits local deployment and makes hosted providers a more practical route for many teams. Watch for independent testing beyond Arena's ranking to see how it performs in real coding work.
VQV organizes public signals from inspectable sources. It does not independently verify the underlying report.
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Scobleizer News Pulse (X).
No login, cookies, social SDKs, or automatic posting.
