WHAT NEW IN X11.06 FRI

AI collections from X · 23 collections · 69,408 posts

jevVIDEO

Jev lands on Vercel AI Gateway to handle agent micro-decisions

Jev is now available on Vercel AI Gateway and can plug into Claude Code, Codex, or browser agents to make tool selection, escalation, and click-routing decisions.

PolyBender
@p0lybenderArchived 2026.09.28Originally posted 2026.09.28
PolyBender @p0lybenderOriginally posted 2026.09.28
Open original on X

Jev is now live on Vercel AI Gateway. plug it directly into Claude Code, Codex, or any AI agent you're already using. 193x faster than traditional LLMs. 444x cheaper. that's not a marketing claim. that's what happens when you build a model that only makes decisions instead of generating text. here's where you actually use it every AI agent has a decision layer. route this to a human or handle it. escalate or respond. click this button or navigate. instead of burning a full LLM call on every micro-decision, you use Jev. Claude Code decides which tool to call next - use Jev. Codex decides whether to execute or ask for clarification - use Jev. your browser agent decides which link to click - use Jev. the math on a real workflow: running ticket triage on 1,000 support requests with Claude: $8. same task with Jev: $0.02. that's the difference between "nice to have" and "actually viable at scale". Vercel AI Gateway means it integrates like every other model in your stack. no new account. no new connection. just another decision engine alongside your reasoning engines. the production shift is quietly happening reasoning models for thinking. decision models for choosing. generative models for creation. the AI stack stops being "use one model for everything" and becomes "use the right model for the layer". Jev fills a gap nobody knew existed until it was 444x cheaper to fill it.

This post quotes another post: x.com/p0lybender/status/2101765571096506841

x.com/p0lybender/status/2104280240277061804Archived 2026.09.28 10:04