obs-ollama-api-docs · inference-research

Fields

id

sources/obs-ollama-api-docs

URL

https://github.com/ollama/ollama/blob/main/docs/api.md

Publisher

ollama

Kind

repo

Accessed

2026-10-04

notes

The per-request timing fields returned in the generate/chat response: total_duration, load_duration, prompt_eval_count, prompt_eval_duration, eval_count, eval_duration, with the docs telling you to compute tok/s as eval_count / eval_duration * 1e9. These are per-response JSON values, not an exported metric series.

Notes

The per-request timing fields returned in the generate/chat response: total_duration, load_duration, prompt_eval_count, prompt_eval_duration, eval_count, eval_duration, with the docs telling you to compute tok/s as eval_count / eval_duration * 1e9. These are per-response JSON values, not an exported metric series.

Cross-references

This record references no other records.