Models · Text

vikasit-235b-moe

Large MoE flagship. Advanced reasoning, agentic workflows, 262K context.

Parameters

235B (22B active)

Context window

262K

Best for

Enterprise AI

Benchmarks

MMLU-Pro84.4%

GPQA Diamond81.1%

AIME 202592.3%

LiveCodeBench74.1%

Thinking mode, no tools where applicable. See full comparisons on the benchmarks page.

# via Vikasit Inference (OpenAI-compatible)

base_url = "https://api.vikasit.ai"

model = "vikasit-235b-moe"