batish52

batish52 / codecontext

Public

Reduce your LLM costs by 40-70% automatically. Routes prompts locally, compacts context, tracks real savings.

pypi.orgprojectpromptrouter anthropic code-search cost-optimization developer-tools gateway
11
1
100% credibility
Found Apr 22, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A drop-in helper that routes simple code questions to local tools, packs tight context for others, and tracks real savings to cut AI bills by 40-70%.

How It Works

1
🔍 Discover a money-saver

You hear about a helpful tool that automatically cuts your AI chat bills for code questions by sending simple ones to a free local helper and trimming extras.

2
📦 Grab it easily

Download and add it to your project folder with a simple command, like adding a helpful app.

3
⚙️ Set it up quick

Create a short settings note in your project and link your AI helper so it knows where to get smart answers.

4
📂 Scan your project

Tell it to look over your files once, building a quick map of what's inside without changing anything.

5
💬 Ask away!

Type a question about your code, like 'where's the login check?', and get a smart reply using just the right bits.

6
📊 Watch savings grow

Check daily reports showing exactly how much money and time it saved you compared to full AI chats.

💰 Bills drop big

Enjoy 40-70% lower AI costs while your code questions get answered faster and smarter.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 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 codecontext?

CodeContext is a Python gateway that slashes LLM costs 40-70% for code-related prompts by routing trivial ones to local models or skipping them, packing only relevant code context via BM25 ranking and embeddings, and logging real token/$ savings to SQLite. Drop it between your app and OpenAI, Anthropic, or Ollama endpoints to automatically compact context from Python/JS/TS repos, reducing llm reduce memory usage and llm reduce inference time without changing your code. CLI commands like `codecontext index-project` and `auto-start-request` get you indexing and querying in under a minute.

Why is it gaining traction?

It stands out with auditable per-request savings reports—no lagged dashboards—and a built-in benchmark to measure your workload's exact reduction in llm cost and llm latency. Minimal deps (just pathspec required), HTTP/Python APIs for easy integration, and support for code context graph tracing across files hook devs tired of bloated prompts causing llm hallucination or high bills. Tracks reduce llm response time via smart compaction, not vague estimates.

Who should use this?

Backend engineers shipping AI code assistants or internal tools hitting OpenAI/Anthropic for repo Q&A/refactors. Teams optimizing github reduce actions cost or llm reduce model size in CI/CD pipelines. Devs evaluating code context ai for precise symbol retrieval without full-repo dumps.

Verdict

Try it if you're burning cash on code LLM calls—CLI and benchmarks make proof-of-concept dead simple. 1.0% credibility and 11 stars flag beta maturity (good docs/tests, but watch for edge cases); production-ready for cost-tracking pilots.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.