llmtop 0.1.0

Realtime TUI monitor for local LLM servers (ollama, llama.cpp). The only GPU monitor that knows what model is running and how much each token costs you in energy and dollar-equivalent.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

use clap::Parser;

#[derive(Parser, Debug, Clone)]
#[command(name = "llmtop", version, about = "Realtime TUI monitor for local LLMs")]
pub struct Cli {
    /// Ollama base URL.
    #[arg(long, default_value = "http://127.0.0.1:11434")]
    pub ollama_url: String,

    /// Compare per-token cost vs this provider model (claude-sonnet | gpt-4o | gemini-2.5).
    #[arg(long, default_value = "claude-sonnet")]
    pub compare: String,

    /// CO2 grid intensity in gCO2eq per kWh (default = world average).
    #[arg(long, default_value_t = 475.0)]
    pub grid_co2: f64,

    /// Disable color output.
    #[arg(long)]
    pub no_color: bool,

    /// Transparent reverse proxy port in front of Ollama. Point your client at
    /// `http://127.0.0.1:<port>` (e.g. `OLLAMA_HOST=...`) to get live tok/s.
    /// Enabled by default; disable with `--no-proxy`.
    #[arg(long, value_name = "PORT", default_value_t = 11435)]
    pub proxy: u16,

    /// Disable the reverse proxy.
    #[arg(long)]
    pub no_proxy: bool,
}

impl Cli {
    pub fn proxy_port(&self) -> Option<u16> {
        if self.no_proxy { None } else { Some(self.proxy) }
    }
}