{"id":30236937,"url":"https://github.com/ihatenodejs/llm-tests","last_synced_at":"2026-02-14T16:32:52.769Z","repository":{"id":309245805,"uuid":"1035548164","full_name":"ihatenodejs/llm-tests","owner":"ihatenodejs","description":"My personal, web-dev focused LLM tests","archived":false,"fork":false,"pushed_at":"2025-08-10T18:43:01.000Z","size":63,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-01-14T01:21:33.292Z","etag":null,"topics":["llm","llm-testing"],"latest_commit_sha":null,"homepage":"","language":"HTML","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"cc0-1.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ihatenodejs.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-08-10T16:23:39.000Z","updated_at":"2025-08-10T18:43:05.000Z","dependencies_parsed_at":"2025-08-10T20:36:20.416Z","dependency_job_id":"b7b05586-e29d-423e-8f50-e724237da78c","html_url":"https://github.com/ihatenodejs/llm-tests","commit_stats":null,"previous_names":["ihatenodejs/llm-tests"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/ihatenodejs/llm-tests","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ihatenodejs%2Fllm-tests","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ihatenodejs%2Fllm-tests/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ihatenodejs%2Fllm-tests/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ihatenodejs%2Fllm-tests/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ihatenodejs","download_url":"https://codeload.github.com/ihatenodejs/llm-tests/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ihatenodejs%2Fllm-tests/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29449370,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-14T15:52:44.973Z","status":"ssl_error","status_checked_at":"2026-02-14T15:52:11.208Z","response_time":53,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["llm","llm-testing"],"created_at":"2025-08-15T01:32:12.758Z","updated_at":"2026-02-14T16:32:52.752Z","avatar_url":"https://github.com/ihatenodejs.png","language":"HTML","funding_links":[],"categories":[],"sub_categories":[],"readme":"# llm-tests\n\nMy personal, web-dev focused LLM tests\n\n## How tests are performed\n\nI do all testing on my MacBook Pro M4, using LM Studio. I am only able to test models up to 64GB, due to memory limits. The setup may change in the future. I aim to create a consistent testing environment, by doing the following:\n\n- Max context length is always used.\n- The same prompt is used for every model.\n- Prompts are written to be **unique**, and different from the samples provided by AI companies.\n- If the LLM provided a structure, it is always followed. No files are combined/split to follow a certain format. Instructions (finding images/manual work) for humans are not followed.\n- I never ask the model to fix it's work. If it's broken, it's used anyway. These tests are performed in one shot. If the model refuses or the output is not a website, the prompt may be re-tried.\n\n## The prompts\n\nFor testing, I prefer to write short, unique, and entertaining prompts. This means the model is not given explicit instructions or help. I believe this makes the outputs more interesting!\n\nI also do not prefer using LLMs for writing. This means the prompts are never AI generated (including this README!), which I believe yields better results.\n\n## How I judge outputs\n\nIn `/test-dir/rank.md`, you will notice I may have ranked the models. But, ranking is subjective. Here's how I judge:\n\n- Did the model fulfill all my requirements? Did it refuse all/some of the prompt?\n- If I was to visit the website as an actual user, would I understand/get everything I needed?\n- I prefer seeing results that are *complete*. This means a header, content, and a footer, when appropriate.\n\nThe ranking files are purely my own opinion, and no points or scoring is done.\n\n## Test directory structure\n\n```\ntest-name/\n  model-name/       # May contain sub directories for things like reasoning effort\n    raw.txt         # The raw output from the model\n    test.xml        # Test environment details\n    content/        # The root directory for the LLM's outputted project\n  [...]\n```\n\n## The inspiration\n\nThis repo is inspired by [Bijan Bowen](https://www.youtube.com/@Bijanbowen)'s videos, along with prompt ideas from his videos, my own imagination, and articles I've read.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fihatenodejs%2Fllm-tests","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fihatenodejs%2Fllm-tests","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fihatenodejs%2Fllm-tests/lists"}