ResearchTasksTime to First Token (TTFT) measurementFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedContext sequences 32K-256K vLLM 0.18.0 / 0.20.0 (online-serving)Qwen3.5-9B (distilled indexer)2.944TTFT (Dense, s)7Jun 9, 2026
Context sequences 32K-256K vLLM 0.18.0 / 0.20.0 (online-serving)Qwen3.5-9B (distilled indexer)2.944TTFT (Dense, s)7Jun 9, 2026