# LucenaCoder adds support for any OpenAI-compatible endpoint

URL: https://press.bz/@lucenaone/lucenacoder-adds-support-for-any-openai-compatible-endpoint
Company: LucenaOne
Date: Jun 8, 2026
Tag: Product

[LucenaOne press office](https://press.bz/@lucenaone)

## Summary

LucenaCoder now routes to any OpenAI-compatible API—self-hosted, regional, or specialty providers—while keeping token efficiency baked into the harness so you pay far less per real task.

## Article

LucenaOne today announced that **LucenaCoder** now supports **any OpenAI-compatible endpoint**, extending the harness beyond its OpenRouter integration.

Teams can point LucenaCoder at the models they actually want—on-prem stacks, regional hosts, or specialty providers—without surrendering to wrappers that dump entire codebases into context just to change one line.

LucenaOne frames the update as the best of both worlds: **direct access to the models you choose**, while **paying much less** because token efficiency is the product. Scoped context, folder-level access, and deterministic workflows stay in the harness; only the endpoint becomes yours to pick.

The release builds on the LucenaCoder beta LucenaOne shipped in May: same client-side defaults, same local API keys, same `npx lucenacoder` local mode—now with a wider pipe for where those tokens go.

LucenaCoder remains free to use. Details at [lucenacoder.com](https://lucenacoder.com).

## Press releases

### [LucenaCoder adds native Codebase.Observer integration for deterministic codebase analysis](https://press.bz/@lucenaone/lucenacoder-adds-native-codebaseobserver-integration-for-deterministic-codebase-)
- Date: Aug 24, 2026
- Tag: Product
- Summary: LucenaCoder now includes native support for Codebase.Observer, bringing deep deterministic codebase analysis to PRO users. Powered by the Evōk Semantic Engine, the integration generates complete architectural blueprints without heuristics or LLMs.

### [LucenaCoder completes multi-file coding task using half the tokens of nearest competitor in Almedra benchmark](https://press.bz/@lucenaone/lucenacoder-completes-multi-file-coding-task-using-half-the-tokens-of-nearest-co)
- Date: Aug 24, 2026
- Tag: Research
- Summary: In the Almedra token efficiency benchmark, LucenaCoder completed the same multi-file task using 61,506 tokens, less than half the 117,002 tokens used by the next closest tool and nearly nine times fewer than the most expensive competitor.

### [LucenaCoder adds side chats for both free and pro users](https://press.bz/@lucenaone/lucenacoder-adds-side-chats-for-both-free-and-pro-users)
- Date: Aug 6, 2026
- Tag: Product
- Summary: LucenaCoder introduces side chats: independent, project-bound conversations available picture-in-picture on Mac, Windows, and Linux, letting developers run parallel coding sessions while keeping token usage scoped and low.

### [LucenaOne releases token-efficiency obsessed agentic coding harness, LucenaCoder](https://press.bz/@lucenaone/lucenaone-releases-token-efficiency-obsessed-agentic-coding-harness-lucenacoder)
- Date: May 16, 2026
- Tag: Beta
- Summary: LucenaCoder enters public beta as a free, privacy-friendly coding harness that runs client-side—built for developers who still have to care what the API bill says.

---

© 2026 LucenaOne
https://press.bz/@lucenaone
- LucenaCoder: [LucenaCoder](https://lucenacoder.com)
- GitHub: [GitHub](https://github.com/lucenaone)