An AI playground for fast prompt iteration.

Point curlgeco at any Hugging Face or OpenAI-compatible endpoint, then stream chat, run prompt suites, and replay every request. Your endpoints, keys, and history stay in your browser.

Open chat lab Configure endpoints

Connections

Save Hugging Face router, dedicated, or generic OpenAI-compatible endpoints.

Open

Chat Lab

Stream completions, tune system prompts, and export sessions.

Open

Test Runs

Benchmark prompt packs, compare endpoints, and share markdown reports.

Open

Replay Logs

Inspect every request, latency pattern, and failure replay path.

Open

curlgeco by ugeco

Open source, and there is no curlgeco server: the app is a static site, so your keys go straight from your browser to the endpoint you configured and nowhere else.

View on GitHub

Workspace snapshot

What curlgeco is carrying right now.

0 endpoints

0 test results

0 log entries

Recent activity

Latest tests and replayable logs.

No activity yet.