LetTokenBurn
Sign in
Auto86%Harbor adapterKnowledge & Reasoning

web-bench

Web-Bench: Unified Web Search QA Benchmarks A curated collection of 7 web search and question-answering benchmarks in a unified format, totaling 3943 questions across 5 base datasets with reference URL variants.

PaperDataset

How to run

ltb · web-bench
$ npm i -g lettokenburn-cli
$ ltb bench pull web-bench --n 50
$ ltb bench run ./web-bench -a claude-code -m anthropic/claude-opus-4-8
$ ltb bench push

Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs

Scores

0 sourced entries

No sourced scores recorded yet.

Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.

Reproduced by the community

All runs
Loading community runs…

About

Web-Bench: Unified Web Search QA Benchmarks

A curated collection of 7 web search and question-answering benchmarks in a unified format, totaling 3943 questions across 5 base datasets with reference URL variants.

📊 Dataset Overview

Dataset Split Records Language URLs Included Description

HLE hle 2,158 EN ❌ Multiple choice questions from Humanity's Last Exam

GAIA-text gaia_text 103 EN ❌ Text-only GAIA benchmark (dev split)

WebWalkerQA webwalkerqa 680 EN ❌…