20-year-old solo builder from Bihar releases open 5.84B omni model, claims it beats Apple AFM 3B on MATH-500 74.2% vs 48.0%
That-Bookkeeper-8316 · reddit · 2026-09-11
Abhinav Anand, a 20-year-old from Bihar who was previously accused of vaporware, released Arcle V1: an open-weight 5.84B unified omni foundation model with a single forward pass across text, vision, documents, and speech, a 2M-token context window, and full offline operation. Self-reported benchmarks vs Apple AFM 3B: MATH-500 74.2% vs 48.0%, BBH 53.7% vs 44.2%, GSM8K 77.5% vs 75.0%. Compute was provided by Lightning AI. Claims are unverified by third parties.
More from Models
- Dev Claims DeepSeek v4.1 Flash "Way Better" Than Gemini Flash Models — gaganghotra_ · 2026-09-12
- Altman Talks GPT7 in New Video Clip — Popular_Slip_5311 · 2026-09-12
- LinkedIn user claims GPT-6 built a pixel-perfect Figma design system in 3 hours — AIandDesign · 2026-09-12
- Report: Google Achieved Recursive Self-Improvement in Next-Gen AI Model, Launching Oct 5 — Dr_Singularity · 2026-09-12
- Leaked Codenames Suggest OpenAI Model Tiers Map to Claude's Opus/Sonnet/Haiku — haider1 · 2026-09-12
- Eric Xing on "open source": open weights is a house with no blueprints — YiMaTweets · 2026-09-12