SOURCE-LINKED INTELLIGENCE
gpt-oss-120b — Server — GB300-NVL72_GB300-288GB_aarch64x8_TRT
127921 Tokens/s; MLPerf v6.1, closed.
Read original source ↗ Open in workspace
- recordType
- benchmark
- region
- Global
- value
- 127921
- unit
- Tokens/s
- dimensions
- {"division":"closed","location":"./closed/Nebius/results/GB300-NVL72_GB300-288GB_aarch64x8_TRT/gpt-oss-120b/Server/performance/run_1","metric":"Performance_Result","model":"gpt-oss-120b","precision":"fp4","release":"v6.1","scenario":"Server","submissionId":null,"submitter":"Nebius","suite":"datacenter","system":"GB300-NVL72_GB300-288GB_aarch64x8_TRT","unit":"Tokens/s"}
Evidence & attribution
MLCommons, MLPerf Inference v6.1, published submission results. Metadata adapted; scenarios and units remain separate.
License: Apache-2.0
First collected: 2026-09-19T22:50:59.123Z. This is not the publication date.