Apertus 70B Instruct
Creator
Swiss AI Initiative
Released
2025-09-02
Intelligence
7.7
Artificial Analysis Index
Coding
1.9
Artificial Analysis Index
In $/1M
$0.82
input tokens
Out $/1M
$2.92
output tokens
Blended $/1M
$1.34
3:1 blended
Speed
0
tokens / sec
Benchmark breakdown
Independent evaluation scores, normalized to 0–100.
GPQA Diamond
27
Humanity’s Last Exam
6
MMLU-Pro—
SciCode
6
LiveCodeBench—
MATH-500—
AIME 2025—
τ²-Bench (agentic)
13
Terminal-Bench Hard
0
IFBench
26