https://github.com/squat/llama-swap-exporter
A Prometheus exporter for llama-swap and the underlying llama.cpp model servers, providing metrics for model state, token usage, etc.
https://github.com/squat/llama-swap-exporter
Last synced: about 1 month ago
JSON representation
A Prometheus exporter for llama-swap and the underlying llama.cpp model servers, providing metrics for model state, token usage, etc.
- Host: GitHub
- URL: https://github.com/squat/llama-swap-exporter
- Owner: squat
- License: apache-2.0
- Created: 2026-05-08T10:33:21.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2026-05-10T23:45:33.000Z (about 2 months ago)
- Last Synced: 2026-05-11T01:31:30.874Z (about 2 months ago)
- Language: Nix
- Size: 1.96 MB
- Stars: 0
- Watchers: 0
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Prometheus llama-swap Exporter
A Prometheus exporter for [llama-swap](https://github.com/mostlygeek/llama-swap) and the underlying [llama.cpp](https://github.com/ggml-org/llama.cpp) model servers, providing metrics for model state, token usage, etc.
[](https://github.com/squat/llama-swap-exporter/actions?query=workflow%3ACI)
[](https://goreportcard.com/report/github.com/squat/llama-swap-exporter)
[](https://builtwithnix.org)
## Prometheus Configuration
Add the following to your `prometheus.yml`:
```yaml
scrape_configs:
- job_name: llama-swap-exporter
static_configs:
- targets: [localhost:9293]
scrape_interval: 30s
metrics_path: /metrics
```
## Metrics
You can find the full list of metrics in the [METRICS.md](./docs/METRICS.md) file.
## Usage
[embedmd]:# (help.txt)
```txt
Usage of llama-swap-exporter:
-api-key string
Bearer token to authenticate against llama-swap; can be provided with the LLAMA_SWAP_EXPORTER_API_KEY environment variable
-metrics-path string
HTTP path on which to serve metrics (default "/metrics")
-scrape.timeout duration
Per-target scrape timeout (default 10s)
-upstream string
Comma-separated llama-swap base URLs
-version
Print the version of llama-swap-exporter and exit
-web.listen-address string
Address on which to serve metrics (default ":9293")
```