LetTokenBurn
Sign in
Verified86%Harbor: pendingMultimodalKnowledge & Reasoning

GDP.pdf

It was cited in Anthropic's Fable 5 and Mythos 5 model card.

PaperDatasetLeaderboard

How to run

ltb · gdp-pdf
$ npm i -g lettokenburn-cli
$ ltb bench info gdp-pdf

External harness. No Harbor adapter has been generated for this benchmark yet. `ltb bench info` prints the known links.

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

Scores

42 sourced entries
HarnessSourceConfidenceLink
GPT-6.1 Sol35.0N/AIndependentmediumN/Asurgehq.ai
GPT-6 Astra34.2N/AIndependentmediumN/Asurgehq.ai
Claude Sonnet 5.531.4N/AIndependentmediumN/Asurgehq.ai
Muse Wonder 1.030.8N/AIndependentmediumN/Asurgehq.ai
GPT-5.6 Sol30.7N/AIndependentmediumN/Asurgehq.ai
Claude Opus 5.530.6N/AIndependentmediumN/Asurgehq.ai
Claude Fable 530.0N/AIndependentmediumN/Asurgehq.ai
Claude Fable 5.129.6N/AIndependentmediumN/Asurgehq.ai
Gemini 4 Argon28.4N/AIndependentmediumN/Asurgehq.ai
Muse Spark 1.327.6N/AIndependentmediumN/Asurgehq.ai
GPT-6 Sol26.4N/AIndependentmediumN/Asurgehq.ai
GPT-5.526.0N/AIndependentmediumN/Asurgehq.ai
GPT-5.6-terra24.7N/AIndependentmediumN/Asurgehq.ai
Claude Opus 4.824.0N/AIndependentmediumN/Asurgehq.ai
Claude Opus 524.0N/AIndependentmediumN/Asurgehq.ai
Gemini 3.7 Flash23.8N/AIndependentmediumN/Asurgehq.ai
Gemini 3.8 Flash23.4N/AIndependentmediumN/Asurgehq.ai
Qwen3.8 Max23.2N/AIndependentmediumN/Asurgehq.ai
GPT-6 Luna23.0N/AIndependentmediumN/Asurgehq.ai
Grok-4.722.8N/AIndependentmediumN/Asurgehq.ai
GPT-5.6 Luna22.7N/AIndependentmediumN/Asurgehq.ai
Claude Opus 4.721.0N/AIndependentmediumN/Asurgehq.ai
DeepSeek V4.1 Flash19.8N/AIndependentmediumN/Asurgehq.ai
Kimi K319.0N/AIndependentmediumN/Asurgehq.ai
Claude Sonnet 4.618.0N/AIndependentmediumN/Asurgehq.ai
Grok-4.617.2N/AIndependentmediumN/Asurgehq.ai
Gemini 3.1 Pro17.0N/AIndependentmediumN/Asurgehq.ai
Qwen3.8 Flash16.6N/AIndependentmediumN/Asurgehq.ai
Muse Spark 1.216.0N/AIndependentmediumN/Asurgehq.ai
Muse Spark 1.115.0N/AIndependentmediumN/Asurgehq.ai
Gemini-3.5-flash14.0N/AIndependentmediumN/Asurgehq.ai
Gemini 3.6 Flash14.0N/AIndependentmediumN/Asurgehq.ai
GLM-5.3 Flash14.0N/AIndependentmediumN/Asurgehq.ai
Grok 4.514.0N/AIndependentmediumN/Asurgehq.ai
Kimi K2.612.0N/AIndependentmediumN/Asurgehq.ai
Gemini 3.5 Flash Lite10.0N/AIndependentmediumN/Asurgehq.ai
Gemini 3 Flash Preview10.0N/AIndependentmediumN/Asurgehq.ai
Grok-4.38.0N/AIndependentmediumN/Asurgehq.ai
Mistral Large 32.0N/AIndependentmediumN/Asurgehq.ai
Nova 22.0N/AIndependentmediumN/Asurgehq.ai
Nova 2.0 Pro Preview2.0N/AIndependentmediumN/Asurgehq.ai
NVIDIA Nemotron 3 Nano Omni2.0N/AIndependentmediumN/Asurgehq.ai

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

GDP.pdf

GDP.pdf measures professional multimodal reasoning over the documents the economy actually runs on: dense reports, contracts, filings, and records in the messy real-world formats professionals work from. It was cited in Anthropic's Fable 5 and Mythos 5 model card.

What it tests

The benchmark contains 100 real-world prompts and PDFs pulled directly from professional workflows across ten domains: Finance, Healthcare, Legal, STEM/Research, Engineering…