Baseten Loops goes GA with serverless SFT/RL training, per-token pricing
baseten · x · 2026-11-08
Baseten announced Loops is now GA with a serverless option. Teams at Ambience AI, Datadog, and LangChain use the Loops SDK for SFT and RL training. Serverless Loops features always-warm infra (no cold starts), pay-per-token pricing for both training and sampling, one-click deployment of trained checkpoints to production inference, and model-specific numerical validation to reduce train/sampling mismatch. Users bring data, reward functions, and training logic; Baseten handles infra and users keep full ownership of weights. A customer notes it enables a tight research feedback loop for continuous model improvement on massive data.
Related event: Baseten Loops Reaches GA with Serverless Post-Training(4 posts)→
More from coding & agent
- Local LLM manager fix: concurrent requests from one session yield 2-3X speedup — draginol · 2026-11-09
- Dev: with latest-gen LLMs he stops overspecifying and lets the model riff — jongranskog · 2026-11-09
- Proof of Agent grades agents with other agents — repetitive bot farms earn nothing — adamamcbride · 2026-11-09
- Hugging Face Dev Asks How to Keep Persistent SSH with Claude Mobile — NielsRogge · 2026-11-08
- DHH: Agents Bring Back the Commodore 64 Era of Computing Invitation — mark_k · 2026-11-08
- Two Anthropic Engineers Reveal Claude Code's Hidden Superpowers in 24 Minutes — hey_abusiddik · 2026-11-08