Ex-Microsoft exec impressed by mysterious Argon model: strong math, tiny context
ivan_bezdomny · x · 2026-10-21
Former Microsoft ad chief Mikhail Parakhin shares early impressions of the mysterious new model Argon, quoting his own earlier take that now seems validated ("Sounds like I won't have to eat my hat after all"). His verdict: Argon's bones are very good — math is strong and post-training is top-notch — but it's hampered by extremely low reasoning effort and a very small context window. He's even publicly asking if anyone knows a hidden Antigravity config parameter to increase context length. The model's origin remains unconfirmed; treat early takes with caution.
More from Models
- One Week After Gemini 4 Argon Launch, Users Are Building 3D Worlds, Physics Sims and Games — minchoi · 2026-10-21
- Minimax H3 local workflow is '0/10': hyper-sensitive prompts and VRAM failures on 24GB — Ambitious_Corgi5723 · 2026-10-21
- ChatGPT's new UI keeps attaching wrong photos to real people, user reports — ProfessionalAd1198 · 2026-10-21
- Tone-filtering AI is "preemptive shame applied at machine speed" — PierceLilholt · 2026-10-21
- Was Opus quietly nerfed? Same prompt, 20 days apart, noticeably worse output — JiliJeanlouis · 2026-10-21
- Users hit a new Claude friction: paternalistic redaction of 'sensitive' data, no explanation — mattbeane · 2026-10-21