IFEvalifeval · chat
87.8%
ReportedOct 2026
Benchmark results
See the tests behind each score. They use the benchmark’s original scoring scale.
Chat#922.81 ratingCommunication#1116.34 ratingCreativity#127.75 ratingWriting#127.75 ratingEconomics#717.72 ratingFinance#533.14 ratingHealthcare#1232.07 ratingInstruction Following#1022.53 ratingLanguage#633.70 ratingLegal#633.24 ratingSpatial Reasoning#2716.18 ratingStructured Output#924.69 rating
Chat4 results
Multi-IFmulti-if · chat
80.6%
ReportedOct 2026
TAU-bench Retailtau-bench-retail · chat
67.8%
ReportedOct 2026
Tau2 Airlinetau2-airline · chat
58.0%
ReportedOct 2026
Coding1 result
CFEvalcfeval · code
21.3%
ReportedOct 2026
Communication2 results
Tau2 Retailtau2-retail · communication
71.9%
ReportedOct 2026
Tau2 Telecomtau2-telecom · communication
45.6%
ReportedOct 2026
Creativity1 result
Creative Writing v3creative-writing-v3 · creativity
86.1%
ReportedOct 2026
General intelligence3 results
Arena-Hard v2arena-hard-v2 · general
79.7%
ReportedOct 2026
BFCL-v3bfcl-v3 · general
71.9%
ReportedOct 2026
Includeinclude · general
81.0%
ReportedOct 2026
Language3 results
MMLU-Prommlu-pro · language
84.4%
ReportedOct 2026
MMLU-ProXmmlu-prox · language
81.0%
ReportedOct 2026
MMLU-Reduxmmlu-redux · language
93.8%
ReportedOct 2026
Math4 results
AIME 2025aime-2025 · math
92.3%
ReportedOct 2026
HMMT25hmmt25 · math
83.9%
ReportedOct 2026
LiveBench 20241125livebench-20241125 · math
78.4%
ReportedOct 2026
PolyMATHpolymath · math
60.1%
ReportedOct 2026
Reasoning6 results
GPQAgpqa · reasoning
81.1%
ReportedOct 2026
Humanity's Last Examhumanity's-last-exam · reasoning
18.2%
ReportedOct 2026
LiveCodeBench v6livecodebench-v6 · reasoning
74.1%
ReportedOct 2026
OJBenchojbench · reasoning
32.5%
ReportedOct 2026
SuperGPQAsupergpqa · reasoning
64.9%
ReportedOct 2026
TAU-bench Airlinetau-bench-airline · reasoning
46.0%
ReportedOct 2026
Writing1 result
WritingBenchwritingbench · writing
88.3%
ReportedOct 2026