imperativelabs

imperativelabs / orkestra

Public

Stop overpaying for LLM calls. Orkestra automatically routes every prompt to the cheapest model that can handle it.

14
2
100% credibility
Found Mar 06, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Orkestra is a Python library that intelligently routes prompts to the most cost-effective AI models across multiple providers to minimize expenses without sacrificing quality.

How It Works

1
📰 Discover Orkestra

You hear about a smart tool that saves money on AI chats by picking the right brainpower for each question.

2
📦 Set it up quickly

With one simple command, you add Orkestra to your project and it's ready to go.

3
🔗 Link your AI friends

You connect services like Google, Anthropic, or OpenAI so Orkestra can choose from their thinkers.

4
💬 Start chatting

You ask any question, from simple facts to tough puzzles, and Orkestra handles the rest automatically.

5
🧠 It picks perfectly

Simple asks go to speedy budget options, hard ones to premium powerhouses, saving you cash every time.

6
📊 Watch savings live

A dashboard pops up showing real-time costs, models used, and how much you're saving—up to 80%.

💰 Chat smarter, spend less

Now your AI conversations are efficient, affordable, and just as smart as before.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is orkestra?

Orkestra is a Python library that stops overpaying for LLM calls by routing every prompt to the cheapest model across OpenAI, Anthropic, or Google that can handle it—budget tiers for trivia, premium for tough tasks. Install via pip, wire up API keys with Provider or MultiProvider, and hit chat() or stream_text() for responses with built-in cost, savings, and token breakdowns. A bundled TypeScript dashboard tracks usage live, plus OpenClaw integration turns it into a drop-in proxy.

Why is it gaining traction?

No-config KNN routing classifies prompts in real-time, delivering 70-80% savings on mixed workloads without quality dips, plus strategies like "cheapest" or "balanced" across providers. Events and middleware let you hook in logging or filters globally or per-provider, with full transparency on routing scores and latency. The self-contained monitor UI and OpenAI-compatible proxy make observability instant, outpacing basic rate-limiters or manual model picks.

Who should use this?

AI engineers scaling agent workflows or apps with variable prompt complexity, tired of blasting GPT-4o at everything. Teams integrating LLMs into tools like OpenClaw, wanting to slash bills without babysitting models. Devs prototyping cost-aware chains before production.

Verdict

Early alpha with 14 stars and 1.0% credibility—docs shine but expect rough edges like auto-downloaded routers. Worth a spin for LLM-heavy prototypes to stop overpaying; production needs more battle-testing.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.