benchgap
Other · model

ERNIE 5.5 Flash Preview benchmark scores

As of 2026-11-18, ERNIE 5.5 Flash Preview (Other) has measured scores on 14 benchmarks and estimated scores on 78 more.

BenchmarkScoreSource
AA-GPQA Diamond91.5%estimated ± 2.7 pp, high confidence
AA-HLE45.2%estimated ± 4.6 pp, high confidence
AA-SciCode54.4%estimated ± 3.1 pp, high confidence
Artificial Analysis Intelligence Index37.3%estimated ± 5.7 pp, medium confidence
CritPt20.3%estimated ± 5.5 pp, medium confidence
GDPval-AA57.4%estimated ± 4.6 pp, medium confidence
GPQA Diamond94.1%estimated ± 2.0 pp, high confidence
HLE54.2%estimated ± 6.5 pp, medium confidence
LiveCodeBench v696.1%estimated ± 1.9 pp, medium confidence
AA-Omniscience Accuracy45.5%estimated ± 7.8 pp, medium confidence
SciCode60.1%estimated ± 7.6 pp, low confidence
BrowseComp87.4%measured
HLE w/ tools59.8%estimated ± 5.6 pp, medium confidence
VITA-Bench46.7%estimated ± 6.7 pp, low confidence
τ²-bench results93.5%estimated ± 11.8 pp, low confidence
AA Coding Index76.5%estimated ± 2.5 pp, high confidence
APEX-Agents64.2%measured
APEX-Agents-AA36.8%estimated ± 4.4 pp, high confidence
DeepSearchQA90.2%estimated ± 7.0 pp, low confidence
SWE-bench Verified82.2%estimated ± 2.3 pp, medium confidence
AutomationBench61.9%measured
BFCL v477.2%estimated ± 4.7 pp, low confidence
CyberGym89.3%estimated ± 4.6 pp, medium confidence
JobBench58.3%estimated ± 8.0 pp, medium confidence
SkillsBench59.9%estimated ± 4.1 pp, medium confidence
SWE-bench Pro62.6%estimated ± 3.5 pp, medium confidence
WideResearch81.2%estimated ± 3.5 pp, high confidence
AA Agentic Index51.8%estimated ± 3.3 pp, medium confidence
MMLU-Pro85.8%estimated ± 2.5 pp, low confidence
GPQA93.5%estimated ± 3.3 pp, high confidence
Vals GPQA Diamond91.6%estimated ± 2.9 pp, high confidence
Vals LiveCodeBench77.4%estimated ± 6.1 pp, low confidence
Vals MMLU-Pro89.2%estimated ± 1.8 pp, high confidence
Chess Puzzles56.6%estimated ± 8.0 pp, low confidence
Gert Labs73.0%estimated ± 7.1 pp, low confidence
AA-AnalystAgent45.0%estimated ± 8.0 pp, low confidence
AA Harvey LAB v1.092.3%estimated ± 9.8 pp, low confidence
AA Tau3 Banking52.0%estimated ± 3.3 pp, low confidence
ApprenticeBench25.5%estimated ± 11.0 pp, low confidence
ARC-AGI-195.2%estimated ± 3.9 pp, high confidence
ARC-AGI-277.7%estimated ± 13.0 pp, low confidence
CursorBench 3.230.2%estimated ± 3.6 pp, low confidence
FrontierCode 1.1 Main50.4%estimated ± 3.5 pp, low confidence
MazeBench3.9%estimated ± 4.0 pp, high confidence
Mystery Game Puzzles61.0%estimated ± 12.6 pp, low confidence
NYT Connections88.2%estimated ± 6.4 pp, low confidence
OSWorld-Verified81.6%estimated ± 5.7 pp, medium confidence
Terminal-Bench 4.145.4%estimated ± 5.0 pp, medium confidence
Vals SWE-bench98.4%estimated ± 5.2 pp, low confidence
CWE-Bench-AA68.6%estimated ± 6.0 pp, low confidence
AA Cyber Index46.3%estimated ± 9.4 pp, low confidence
DeepsecBench-AA20.2%estimated ± 5.2 pp, low confidence
HealthBench Professional (length-adjusted, Artificial Analysis)58.1%estimated ± 4.4 pp, medium confidence
HealthBench Professional (raw, Artificial Analysis)65.9%estimated ± 2.6 pp, medium confidence
AA ITBench49.0%estimated ± 4.1 pp, high confidence
Bug Hunt Bench32.0%estimated ± 7.2 pp, low confidence
CursorBench 4.047.4%estimated ± 4.7 pp, high confidence
DeepSWE71.2%measured
DRACO84.9%measured
HLE w/o tools46.4%measured
OSWorld 2.019.6%estimated ± 5.5 pp, low confidence
PostTrainBench v1.139.2%estimated ± 4.5 pp, high confidence
ProgramBench29.5%measured
Terminal-Bench 4.046.5%measured
Terminal-Bench-Science 0.121.0%measured
Toolathlon-Verified75.0%measured
VulcanBench Frontier v488.5%estimated ± 1.2 pp, medium confidence
AA Terminal-Bench 4.043.5%estimated ± 6.3 pp, medium confidence
ARC-AGI-37.5%estimated ± 7.2 pp, medium confidence
CWE-bench62.9%estimated ± 11.0 pp, low confidence
MKA37.2%estimated ± 10.6 pp, low confidence
Hermes Bench71.1%estimated ± 6.4 pp, low confidence
Hermes Index42.2%estimated ± 10.2 pp, low confidence
ExploitGym21.0%estimated ± 9.2 pp, low confidence
Claw-Eval73.1%estimated ± 6.3 pp, low confidence
DeepPlanning52.2%estimated ± 13.1 pp, low confidence
MCP Atlas84.6%measured
NL2Repo64.8%measured
QwenClawBench69.8%estimated ± 2.6 pp, medium confidence
Toolathlon57.6%estimated ± 4.2 pp, high confidence
MedXpertQA (Text)63.7%estimated ± 8.9 pp, low confidence
ResearchClawBench19.8%estimated ± 1.8 pp, medium confidence
HealthBench Professional55.1%estimated ± 4.2 pp, medium confidence
ScarfBench67.2%estimated ± 3.8 pp, medium confidence
BlenderWorld (text-only)79.0%measured
Cookie-Bench88.2%measured
Agents' Last Exam36.8%estimated ± 11.6 pp, low confidence
OpenHarmony Bench v1.059.8%estimated ± 2.7 pp, high confidence
SWE-Marathon45.0%estimated ± 12.9 pp, low confidence
LifeVal77.9%estimated ± 1.6 pp, medium confidence
PinchBench90.6%estimated ± 3.6 pp, low confidence
SWE-Atlas Codebase QnA60.9%estimated ± 4.8 pp, high confidence