Models · Text
vikasit-32b
Largest dense model on CPU. Best quality for reasoning and code.
Parameters
32B dense
Context window
256K
Best for
CPU deployment
Benchmarks
GPQA Diamond68.4%
AIME 202481.4%
AIME 202572.9%
LiveCodeBench65.7%
MATH-50097.2%
Thinking mode, no tools where applicable. See full comparisons on the benchmarks page.
Run it
# via Vikasit Inference (OpenAI-compatible)
base_url = "https://api.vikasit.ai"
model = "vikasit-32b"