A test run is a single execution of a benchmark test using a defined model configuration.
Each run represents how a particular large language model (LLM) — such as GPT-4, Claude-3, or Gemini — performed on a given task at a specific time, with specific settings.
A test run includes:
Together, test runs make it possible to compare models, providers, and configurations across benchmarks in a transparent and reproducible way.
newspaper-page data-correction 18 en
anthropic claude-opus-4-6 · temp 0.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"The opening TITLE tag was missing its closing angle bracket. Changed \"<TITLEEine\" to \"<TITLE>Eine\" to properly form the opening <TITLE> tag."}newspaper-page data-correction 18 en
genai gemini-3.1-pro-preview · temp 0.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"Added missing closing angle bracket '>' to the opening TITLE tag."}newspaper-page data-correction 18 en
openai gpt-5.4-2026-03-05 · temp 1.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"Corrected the malformed opening TITLE tag from '<TITLEEine' to '<TITLE>' so the XML is well-formed."}newspaper-page data-correction 18 en
anthropic claude-sonnet-4-6 · temp 0.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"The opening tag `<TITLEEine` was malformed — it was missing the closing `>` after `TITLE`. It has been corrected to `<TITLE>`, properly separating the tag from the text content \"Eine Tossanische Bibel\"."}newspaper-page data-correction 18 en
openai gpt-5.2-2025-12-11 · temp 1.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"Added missing opening/closing tags around the TITLE text: changed <TITLEEine Tossanische Bibel</TITLE> to <TITLE>Eine Tossanische Bibel</TITLE>. No other structural issues detected."}newspaper-page data-correction 18 en
openai gpt-5.1-2025-11-13 · temp 1.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>
<NR>37.</NR>
<BIBL>
<TITLE>Eine Tossanische Bibel</TITLE>
</BIBL>
</ITEM>","number_of_corrections":3,"explanation":"1) Wrapped the leading number and dot (\"37.\") in a <NR> element to separate record numbering from the bibliographic block. 2) Inserted the missing closing angle bracket in <TITLEEine Tossanische Bibel</TITLE> so that it becomes a valid <TITLE> element. 3) Reformatted the XML with line breaks and indentation for clarity; structural content remains the same aside from the added <NR> element."}newspaper-page data-correction 18 en
openai gpt-5-nano-2025-08-07 · temp 1.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"Fixed missing closing angle bracket after TITLE tag in <TITLE> Eine Tossanische Bibel</TITLE>. The corrected XML is well-formed with properly nested elements."}newspaper-page data-correction 18 en
openai o3-2025-04-16 · temp 1.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM><NUM>37.</NUM><BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":2,"explanation":"1. Inserted missing closing angle bracket in the opening TITLE tag (<TITLE>).
2. Wrapped the ordinal \"37.\" inside a <NUM> element to separate metadata from bibliographic content, improving structural clarity."}newspaper-page data-correction 18 en
openai gpt-5-2025-08-07 · temp 1.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":1,"explanation":"Added the missing closing angle bracket in the TITLE start tag to form <TITLE>."}newspaper-page data-correction 18 en
openai gpt-4.1-2025-04-14 · temp 0.0 · dataclass CorrectedAdvert
One model response from this run. The per-input comparison needs JavaScript.
{"fixed_xml":"<ITEM>37. <BIBL><TITLE>Eine Tossanische Bibel</TITLE></BIBL></ITEM>","number_of_corrections":2,"explanation":"1. Added missing '>' after <TITLE to form <TITLE>.
2. Inserted a space between <TITLE> and the book title for readability, though this is optional in XML."}