Elis Token Saver

Elis Token Saver

Reduce LLM costs by routing tasks to smaller models

thenoahsalas
@thenoahsalas
Published on Oct 24, 2026
Visit site
$100k MRR
1 PeerPush
🚀
Awarded
Just Launched
PeerPush
MCP server https://api.tryelisai.com/api/mcp/

Details

Follow on
Platforms

Discovery signals

How AI and people discover Elis Token Saver on PeerPush

AI-readyAI-readyWhether this product carries the structured data - use cases, audiences, platforms - that lets AI match it to the right questions.
Structured

Described for AI to match

Described for AI with use cases, audiences and platforms so assistants can match it to the right questions.
Discoverable nowDiscoverable nowWhether this product is queryable through the PeerPush API and MCP right now.
Live

Via the PeerPush API and MCP

Queryable through the PeerPush API, MCP and semantic search from day one.

About Elis Token Saver

**Elis Token Saver — MCP** *Stop paying for tokens you didn't need to spend.* Elis Token Saver is an MCP server that plugs into the coding agent or chat host you already use — Claude Code, Codex, Gemini CLI, Copilot — and turns one large, expensive request into a right-sized execution plan. Instead of letting a frontier model chew through an entire task end to end, Elis decomposes the goal into discrete tasks, assigns each one the smallest model that can actually do it, and runs independent tasks in parallel. You keep your host, your models, and your subscription. You just stop overspending on them. **How it works** - Send a goal. Elis returns a free quote first: estimated tokens saved, task count, planning cost. No task content, no commitment. - Accept, and Elis discloses the plan — each task with a model-size recommendation, suggested model, tools, and an assigned agent identity. - Tasks with no dependency between them run concurrently. Failures retry, then trigger an automatic replan; work you discover mid-run gets appended without starting over. - Finished plans are stored privately as proven chains, so the next similar goal skips planning and reuses what already worked. **What makes it different** - **You pay on savings, not usage.** The charge is the reserved estimate from planning. Actual execution tokens are recorded for accuracy — they never move the bill. - **Your data stays yours.** Proven chains live in your own org scope. Community sharing is opt-in, redacted, and review-gated; any single plan can opt out. - **It gets cheaper as you use it.** Feedback grades every chain, demotes weak ones, and reinforces the ones that hold up. A couple of things to fill in before this ships: the savings figure you want to lead with (the copy is deliberately claim-free right now), and where this lands — a directory listing needs a tighter 50-word cut, a landing page can carry all of the above. Tell me which and I'll size it.

Reviews (0)

No reviews yet. Be the first to rate this product!

Comments (1)

thenoahsalas
@thenoahsalas

Your coding agent shouldn't re-plan the same work in its own context window every time. Elis does the planning once, server-side, and hands back the tasks. 80% usage savings on average. 1.8x more time