GDP.pdf
100 tasks built from real professional PDFs, such as manuals, wiring diagrams, blueprints and contracts, across ten fields. The model must find the passage, table or diagram that matters and answer from it. Share of tasks that meet every criterion in their rubric.
Understanding photos, diagrams and documents alongside text. One of 4 comparable tests in multimodal, a category the Index averages.
Read with care. A task counts only when the answer meets every criterion in its rubric, so scores are low for every model. Written and graded by Surge AI.
Publisher: Surge AIWhat a model is asked to do
Answer a question from real working documents: find the right page, table or diagram in a PDF and get every required fact right.
For exampleFrom the attached 60-page equipment manual, give the torque for the cover bolts on the 2019 revision and the tool the manual says to use, and nothing the manual does not say.
Why it matters. Most professional knowledge lives in PDFs. An answer that sounds right but is not in the document is the failure that matters at work.
The example is original and illustrative, not an item from the dataset.
- Models scored here
- 37
- Who produced the numbers
- Official leaderboard
- Items graded
- n = 100 (1.0% each)
- Best published result
- 35.0%
- Newest release covered
- October 12, 2026 (active)
- License
- CC BY 4.0 (via Epoch AI)
- Direction
- Higher is better
- Scale
- Published as a share of items solved. For the Index, each score is a share of the best published result (35.0%), so the frontier reads 100. Guessing earns nothing on this test, so no chance correction applies.
- Provenance
- Produced by the benchmark's own leaderboard or a third-party evaluator, not by the model's maker.
Snapshot November 10, 2026
Ranking on this test
37 models on GDP.pdf
Best published run per model, and how close each one comes to the best published result on this test (the frontier reads 100). Every number links to the publisher that produced it. Internal runs are marked and carry their run details on the model page. One item on this test is 1.0 points, so read gaps smaller than that as noise.
The best models, ranked here, working inside Theo.
29 of the models on this page run inside Theo today. Theo picks the right one for each step and always shows which engine answered.