GLM-5.1 is Zhipu's next-level open source model, ranking #1 in open source and #3 globally across SWE-Bench Pro, Terminal-Bench, and NL2Repo (as of April 2026). Built for long-horizon tasks, it can run autonomously for up to 8 hours. Run at FP8.
Added Mar 27, 2026
Context Window
200.0K
Max Output
131.1K
Input Price (Auto)
$0.79/1M
Output Price (Auto)
$2.73/1M
Capabilities
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
51.4
Choose explicit providers for this model. Auto routing remains available as the default option.
Loading provider options…
Coding Index
43.4
Agentic Index
66.0
GPQA Diamond
Graduate-level scientific reasoning
86.8%
Better than 93% of models compared
HLE
Humanity's Last Exam
28.0%
Better than 93% of models compared
IFBench
Instruction-following benchmark
76.3%
Better than 96% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
97.7%
Better than 99% of models compared
AA-LCR
Long context reasoning evaluation
62.3%
Better than 82% of models compared
GDPval-AA
Economically valuable tasks
49.8%
Better than 94% of models compared
CritPt
Research-level physics reasoning
0.0%
Better than 30% of models compared
SciCode
Python programming for scientific computing
43.8%
Better than 90% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
43.2%
Better than 93% of models compared
AA-Omniscience Accuracy
Proportion of correctly answered questions
25.7%
Better than 77% of models compared
AA-Omniscience Hallucination Rate
Rate of incorrect answers among non-correct responses
63.4%
Better than 81% of models compared
Last updated May 11, 2026
Artificial Analysis