AstaBench
Allen Institute for AI (Ai2)
Before trusting an AI research helper with your science work, you want to see how it does on set tasks. This is an open code project that tests research assistants against scientific tasks and publishes scoreboards of the results.
QUICK INFO
SOLUTION TYPE
Models, data & tools
TECHNOLOGY CATEGORY
Benchmarks & evaluations
AREA OF WORK
Research & scientific knowledge
ACCESS
Free
DEPLOYMENT
Self-hosted / on-premise · Downloadable
SOURCE REVIEWED
August 23, 2026
What it helps with
- Which research assistant should i trust
- Compare ai research helpers
- Test an assistant on science tasks
- See scoreboards for research agents
- Check an ai tool before using it for science
Scope & audience
WHO IT'S FOR
Scientist / researcher · Developer / AI builder
AREA OF WORK
Research & scientific knowledge
COMMODITY / SCOPE
Cross-commodity
Access & technical details
ACCESS
Anyone can get it
INTERFACE
Code repository
DEPLOYMENT
You run it yourself
LANGUAGES
Not stated by provider
Featured media
Source information
This listing was compiled from publicly available source material. Verify current details with the provider.
PROVIDER DOCUMENTATION
Reviewed August 2026
Explore more like this
GEO-Bench
ServiceNow Research
Models, data & tools
Before trusting an AI system on a farming or food task you need a way to check it. Earth observation benchmark for pretrained models, including crop type and cashew plantation segmentation tasks, downloaded from Hugging Face.
Free · Open source
Self-hosted / on-premise · Downloadable
Benchmarks & evaluations
GEO-Bench
research-develop
research-and-scientific-knowledge
buildable-or-downloadable-component
benchmarks-and-evals
Free · Open source
Self-hosted / on-premise · Downloadable
August 17, 2026
- Test a satellite image model on crop tasks
- Compare pre trained earth observation models
- Crop type segmentation benchmark
- Cashew plantation mapping test
- Download benchmark data for satellite models
ruTXTAgroBench
Kuban State Agrarian University, published via MERA
Models, data & tools
Before trusting an AI system on a farming or food task you need a way to check it. 2,935 Russian multiple-choice questions on botany, crop production and genetics testing agronomy knowledge acquired in pretraining.
Free
Self-hosted / on-premise · Downloadable
Benchmarks & evaluations
ruTXTAgroBench
research-develop
research-and-scientific-knowledge
buildable-or-downloadable-component
benchmarks-and-evals
Free
Self-hosted / on-premise · Downloadable
August 17, 2026
- Test a model on Russian agronomy questions
- Does a model know botany and crop production
- Russian multiple choice farming questions
- Compare models on Russian agricultural knowledge
- Genetics questions in Russian for model testing
Is this your Resource?
Contact us to claim this listing and add details, screenshots or a demo.