Legacy
DeepSearch
Ask a question and watch it search, read and reason its way to a cited answer.
Classifier
Label text or images against classes you name, with no training data.
Segmenter
Split text into tokens and chunks, and see the count you will be billed for.
Toy examples
format_list_numbered
Listwise Test
Rank a list of documents against one query and compare the ordering across models.
join_full
Pairwise Test
Score how close two texts are, one pair at a time or a CSV at once.
acute
Reader Speed Test
Race the reader engines on a URL and see which returns usable markdown fastest.
category_search
Image Rerank Test
Rank images against a text query to check multimodal retrieval.
image_search
Visual QA Test
Ask a question about an image and read the answer back from the VLM.
document_scanner
Document OCR
Turn a scanned page, photo or PDF into clean Markdown, tables and equations included.
rule
Structured Decisions
Ask a reranker for a named choice or a graded score, and read off the distribution how sure it actually is.
find_in_page
Semantic Highlighter Test
Find the passage inside a long document that answers a question.
Utilities
speed
Rate limit
The request and token limits that apply to your key, per endpoint.
bar_chart
Token Counter
Count one input across every model at once, and see where the tokens go.
memory
Omni Provisioning
Size the GPUs, time and storage for an omni corpus, from a measured sweep.
calculate
Config License
Estimate what a workload costs before committing to it.
conversion_path
Integrate
Vector databases, frameworks and the meta-prompt for wiring Jina into them.