Swapping RLM sub-calls for Jev: half the cost, 6x faster at 1M tokens
lateinteraction · x · 2026-10-16
A weekend experiment replaced Recursive Language Models' sub-calls with Jev, a System One model returning calibrated probabilities. On OOLONG at 1M tokens, RLM+Jev is half the cost and 6x faster than Haiku 5.5 sub-calls while holding scores, while full-context reads get expensive and degrade. The thread also covers abstention behavior, routing to bigger models, and where the setup didn't help.
More from coding & agent
- Turn charts, logos and photos into animated videos in 5 minutes with Claude Code + Remotion — 6 copy-paste prompts — TawohAwa · 2026-10-16
- Enterprise AI agents on 20-year-old healthcare billing data: CPTO went from skeptic to user in 6 weeks — alex_verem · 2026-10-16
- Anthropic eng lead Felix Rieseberg's Claude tutorial distilled into 7 prompts — alex_verem · 2026-10-16
- LangChain Guide: How Madrigal, Abridge and Vizient Scale Agents in Healthcare — LangChain · 2026-10-16
- Full Masterclass: Building an AI GTM Army That Runs Your Sales Pipeline — VibeMarketer_ · 2026-10-16
- Beam Scores 28 on Coding Agent Index, 40 Behind Leaders but Among Most Token-Frugal — ArtificialAnlys · 2026-10-16