# Tool-call Latency on Novel agentic workflow 6-turn

Source: [Tool-call Latency on Novel agentic workflow 6-turn](<https://www.sota2.com/research/sota/tool-call-latency-on-novel-agentic-workflow-6-turn>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-27T08:00:27.668Z

Evaluation of tool-call latency performance for a 6-turn agentic workflow, measured by median per-turn latency and relative latency against a baseline.

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 | Tool-call Latency |
| Dataset (including any stated split) | Novel agentic workflow 6-turn |
| Metric | Median per-turn Latency (ms) |
| Unit | TIME |
| Direction | Lower is better |
| Results loaded | 3 of 3 |
| 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

- [Median per-turn Latency (ms)](<https://www.sota2.com/research/sota/tool-call-latency-on-novel-agentic-workflow-6-turn.md?metric=Median%20per-turn%20Latency%20(ms)>): Lower is better; unit: TIME.

- [Relative Latency (vs LayerScale)](<https://www.sota2.com/research/sota/tool-call-latency-on-novel-agentic-workflow-6-turn.md?metric=Relative%20Latency%20(vs%20LayerScale)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Median per-turn Latency (ms) | Result date | Report paper |
| --- | --- | --- | --- |
| LayerScale | 624 | Not recorded | 6a167aa3dbd7150506f8002d |
| vLLM | 1338 | Not recorded | 6a167aa3dbd7150506f8002d |
| SGLang | 1400 | Not recorded | 6a167aa3dbd7150506f8002d |

### LayerScale

Median per-turn Latency (ms): 624. Result ID: 6a16a3e2fbc373592c0f9316.

[Reporting paper](<https://www.sota2.com/research/paper/6a167aa3dbd7150506f8002d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a167aa3dbd7150506f8002d.md>)

Recorded evaluation configuration: {"caching":"no response-cache hits"}. This record does not assert independent reproduction.

### vLLM

Median per-turn Latency (ms): 1338. Result ID: 6a16a3e2fbc373592c0f9317.

[Reporting paper](<https://www.sota2.com/research/paper/6a167aa3dbd7150506f8002d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa708c1e209131cea4ba94.md>)

Recorded evaluation configuration: {"caching":"no response-cache hits"}. This record does not assert independent reproduction.

### SGLang

Median per-turn Latency (ms): 1400. Result ID: 6a16a3e2fbc373592c0f9318.

[Reporting paper](<https://www.sota2.com/research/paper/6a167aa3dbd7150506f8002d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa738a1e209131cea5d174.md>)

Recorded evaluation configuration: {"caching":"no response-cache hits"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
