JapanFold launches: 8 open biology AI models served free for Japanese researchers
DavidBennett__ · x · 2026-09-08
ai& has launched JapanFold, bundling 8 open biology AI models—covering protein structure prediction, molecular interactions, binder design, and protein search/classification—into one platform free for Japanese researchers, universities, and institutions via browser, unified API, or AI agents, with inference costs borne by ai&.
Key differences from LLM serving
- LLMs are autoregressive and memory-bandwidth-bound, relying on KV caches, Paged Attention, continuous batching, and speculative decoding
- Folding models are compute-bound: each prediction needs massive sequence database searches, amino-acid relation computation, and multiple structure candidates, consuming heavy memory and compute
- Models differ in data and runtime needs, requiring caching, input-based job routing, and efficient resource scheduling
Context: Over 200M predicted structures are public since AlphaFold. ai& calls JapanFold its first step into AI for Science, arguing the inference platform matters as much as the models themselves.
More from Infra
- Running a local MLX music model via Codex pushes M4 MacBook Air to its limits — Dimillian · 2026-09-08
- Qdrant releases 10B-vector retrieval dataset with exact top-1000 ground truth for 100K queries — qdrant_engine · 2026-09-08
- Debunking the viral 'data center infrasound harm' videos: citations don't hold up — AndyMasley · 2026-09-08
- AI cluster builders bypass the grid: gas turbines, fuel cells and SMRs come onsite — AccBalanced · 2026-09-08
- Jensen Huang calls compute a rentable asset as H100 rental prices jump 22% to $3.28/hr — AccBalanced · 2026-09-08
- Inside the VMs Powering Mobile AI Agents: Instinct, Claude Code — RohanAdwankar · 2026-09-08