Valence AI open-sources Nesso-1 for faster protein binding prediction
Francesco_dgv · x · 2026-07-21
Recursion’s Valence AI says it is releasing Nesso-1, an open-source co-folding-based binding affinity model for drug discovery.
Key claims:
- 1 second per prediction, about 20x faster than their previous Boltz-2 collaboration
- Accuracy that matches or exceeds Boltz-2 on public and internal benchmarks
- 2–3x additional speedup in training and inference from NVIDIA cuEquivariance
- Weights and code are fully open source
Related event: Valence AI Open-Sources Nesso-1(3 posts)→
More from Models
- OpenAI rolls out voice in GPT-Live, but the UI obscures search and reasoning — Graham_dePenros · 2026-07-22
- Moonshot’s Kimi K3 sets a new open-weights ECI record at 156 — scaling01 · 2026-07-22
- Nanbeige4.2-3B launches as a 3B Looped Transformer model that beats larger baselines — Wooden-Deer-1276 · 2026-07-22
- A post says six companies now beat Google’s best LLM, including two open-source models — soham_btw · 2026-07-22
- Google says Gemini 3.5 Pro is in testing and Gemini 4 is already pre-training — Wide-Ad1564 · 2026-07-22
- Gemini 3.5 Flash Lite Tested: Not Frontier-Optimal, but Hits 350 tok/s — brandon_galang · 2026-07-22