# LLM Inference on Merge Source: [LLM Inference on Merge]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:46:01.427Z Evaluation of LLM Inference performance on the Merge dataset, specifically measuring throughput using Goodput (requests per second). A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | LLM Inference | | Dataset (including any stated split) | Merge | | Metric | Goodput (req/s) | | Unit | TIME | | Direction | Higher is better | | Results loaded | 18 of 18 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Goodput (req/s)](): Higher is better; unit: TIME. ## Recorded results | Method | Goodput (req/s) | Result date | Report paper | | --- | --- | --- | --- | | AugServe | 1.16 | Not recorded | 693115d6d6b0dd3342933d87 | | AugServe | 1.15 | Not recorded | 693115d6d6b0dd3342933d87 | | AugServe | 0.87 | Not recorded | 693115d6d6b0dd3342933d87 | | AugServe | 0.58 | Not recorded | 693115d6d6b0dd3342933d87 | | AugServe | 0.47 | Not recorded | 693115d6d6b0dd3342933d87 | | AugServe | 0.35 | Not recorded | 693115d6d6b0dd3342933d87 | | InferCept | 0.28 | Not recorded | 693115d6d6b0dd3342933d87 | | vLLM | 0.22 | Not recorded | 693115d6d6b0dd3342933d87 | | InferCept | 0.2 | Not recorded | 693115d6d6b0dd3342933d87 | | vLLM | 0.15 | Not recorded | 693115d6d6b0dd3342933d87 | | InferCept | 0.13 | Not recorded | 693115d6d6b0dd3342933d87 | | InferCept | 0.11 | Not recorded | 693115d6d6b0dd3342933d87 | | vLLM | 0.1 | Not recorded | 693115d6d6b0dd3342933d87 | | InferCept | 0.08 | Not recorded | 693115d6d6b0dd3342933d87 | | vLLM | 0.07 | Not recorded | 693115d6d6b0dd3342933d87 | | InferCept | 0.06 | Not recorded | 693115d6d6b0dd3342933d87 | | vLLM | 0.05 | Not recorded | 693115d6d6b0dd3342933d87 | | vLLM | 0.04 | Not recorded | 693115d6d6b0dd3342933d87 | ### AugServe Goodput (req/s): 1.16. Result ID: 6996ce165698665ec41a1029. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":1,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### AugServe Goodput (req/s): 1.15. Result ID: 6996ce165698665ec41a1027. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":1.5,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### AugServe Goodput (req/s): 0.87. Result ID: 6996ce165698665ec41a1026. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":1,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### AugServe Goodput (req/s): 0.58. Result ID: 6996ce165698665ec41a102a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":1.5,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### AugServe Goodput (req/s): 0.47. Result ID: 6996ce165698665ec41a1028. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":2,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### AugServe Goodput (req/s): 0.35. Result ID: 6996ce165698665ec41a102b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":2,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### InferCept Goodput (req/s): 0.28. Result ID: 6996ce165698665ec41a1020. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":1,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### vLLM Goodput (req/s): 0.22. Result ID: 6996ce165698665ec41a101a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":1,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### InferCept Goodput (req/s): 0.2. Result ID: 6996ce165698665ec41a1023. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":1,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### vLLM Goodput (req/s): 0.15. Result ID: 6996ce165698665ec41a101d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":1,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### InferCept Goodput (req/s): 0.13. Result ID: 6996ce165698665ec41a1021. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":1.5,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### InferCept Goodput (req/s): 0.11. Result ID: 6996ce165698665ec41a1024. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":1.5,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### vLLM Goodput (req/s): 0.1. Result ID: 6996ce165698665ec41a101b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":1.5,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### InferCept Goodput (req/s): 0.08. Result ID: 6996ce165698665ec41a1022. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":2,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### vLLM Goodput (req/s): 0.07. Result ID: 6996ce165698665ec41a101e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":1.5,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### InferCept Goodput (req/s): 0.06. Result ID: 6996ce165698665ec41a1025. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":2,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### vLLM Goodput (req/s): 0.05. Result ID: 6996ce165698665ec41a101c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":2,"CV":2,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. ### vLLM Goodput (req/s): 0.04. Result ID: 6996ce165698665ec41a101f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Request rate":3,"CV":2,"Model":"OPT-13B","GPU":"H800"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.