Flagship Nemotron model for high-throughput reasoning and complex agents

Details

FieldValue
idnvidia/Llama-3_1-Nemotron-Ultra-253B-v1
familynemotron
modalitytext->text
context_length128000
max_input_tokens120000
max_output_tokens4096
quantizationfp8
knowledge_cutoff2024-12
release_date2025-01-15
is_open_weightsyes
supports_tool_callyes
supports_reasoningno
supports_structured_outputyes
supports_attachmentno
input_modalitiestext
output_modalitiestext

Providers