Switching tweet classification to Jev: 6x faster, ~40x cheaper than fastest LLMs
altryne · x · 2026-09-17
Developer @altryne's Chrome extension, Twitter Timeline Analyzer, uses AI to analyze and filter your Twitter/X "For You" feed to check whether it's obsessed with a single topic. He just integrated TypeSafe's Jev decision engine (with @diogoalmeida).
Benchmarks: previously he used the fastest LLMs on Cerebras; Jev is 6x faster and 40x cheaper on this task — per-tweet yes/no calls in 150–300ms at about $0.00002 each.
Design: your LLM writes each topic's criteria once, then Jev judges every tweet against them. The extension works with any OpenAI-compatible API (Cerebras, Groq, OpenAI, Together, OpenRouter, etc.), can tag, highlight, or hide tweets, learns from manual feedback, and ships with W&B Weave observability. Open source on GitHub.
Related event: Jev Classifies Tweets 6x Faster and 40x Cheaper than Fastest LLM(2 posts)→
More from coding & agent
- Claude Code tip: omitClaudeMd: true stops subagents from burning tokens on CLAUDE.md — daniel_mac8 · 2026-09-18
- Sifting 1,000 X posts via pairwise LLM ranking yields a Claude Code tip: omitClaudeMd — daniel_mac8 · 2026-09-18
- Zach Lloyd: A humming software factory should be a closed-loop, measurable system — vikvang1 · 2026-09-18
- Grok Build ships cross-session memory that logs conventions and decisions to Markdown — minchoi · 2026-09-18
- Grok Build adds cross-session memory: project conventions auto-written to markdown — minchoi · 2026-09-18
- Six real-world uses for TypeSafe's Jev judgment model, 4x faster than Gemini in evals — HamelHusain · 2026-09-18