~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Everything Local & Encrypted.
Python academiaanthropicarxivbravedeep-researchencryptionhome-automationhomeserverlocallocal-deep-researchlocal-llmmistralollamaopenaipubmedresearchresearch-toolretrieval-augmented-generationsearxngself-hosted
fix(templates): drop maintainer's local path from served JS comment (#5878) (#5890)
LearningCircuitbf8639af3484
diff --git a/src/local_deep_research/web/templates/pages/benchmark_results.html b/src/local_deep_research/web/templates/pages/benchmark_results.htmlindex 26adb7e004..95be2615e5 100644--- a/src/local_deep_research/web/templates/pages/benchmark_results.html+++ b/src/local_deep_research/web/templates/pages/benchmark_results.html@@ -634,7 +634,7 @@ async function downloadBenchmarkYAML(runId, includeExamples = false, includeSett// YAML's `date_tested` field MUST come from the benchmark's actual// start_time (or created_at as fallback). External parsers — most- // importantly /home/coder1/code/ldr-benchmarks/scripts/build_leaderboards.py+ // importantly ldr-benchmarks/scripts/build_leaderboards.py// — read this field for the public HuggingFace leaderboard. Pulling// from new Date() (the old behaviour) silently corrupted that history.const startedAt = meta.started_at || run.start_time || run.created_at;