paceflow 0.2.4

Local-first CLI that turns AI coding session history and git metadata into engineering analytics.
Documentation 
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
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156

#[derive(Debug, Clone, Copy, PartialEq)]
pub struct TokenUsage {
    pub input_tokens: i64,
    pub cached_input_tokens: i64,
    pub cache_creation_tokens: i64,
    pub output_tokens: i64,
    pub reasoning_tokens: i64,
}

#[derive(Debug, Clone, Copy, PartialEq)]
pub struct ModelPricing {
    pub input_per_million: f64,
    pub cached_input_per_million: f64,
    pub cache_creation_per_million: f64,
    pub output_per_million: f64,
}

pub fn estimate_cost_usd(model_name: &str, usage: TokenUsage) -> Option<f64> {
    let pricing = pricing_for_model(model_name)?;
    Some(cost_from_pricing(pricing, usage))
}

pub fn cost_from_pricing(pricing: ModelPricing, usage: TokenUsage) -> f64 {
    let billable_output_tokens = usage.output_tokens.saturating_add(usage.reasoning_tokens);
    (usage.input_tokens.max(0) as f64 * pricing.input_per_million
        + usage.cached_input_tokens.max(0) as f64 * pricing.cached_input_per_million
        + usage.cache_creation_tokens.max(0) as f64 * pricing.cache_creation_per_million
        + billable_output_tokens.max(0) as f64 * pricing.output_per_million)
        / 1_000_000.0
}

pub fn pricing_for_model(model_name: &str) -> Option<ModelPricing> {
    let normalized = normalize_model_name(model_name);
    if normalized.contains("gpt-5.5") {
        return Some(ModelPricing {
            input_per_million: 5.0,
            cached_input_per_million: 0.5,
            cache_creation_per_million: 5.0,
            output_per_million: 30.0,
        });
    }
    if normalized.contains("gpt-5.4-mini") {
        return Some(ModelPricing {
            input_per_million: 0.75,
            cached_input_per_million: 0.075,
            cache_creation_per_million: 0.75,
            output_per_million: 4.5,
        });
    }
    if normalized.contains("gpt-5.4") || normalized.contains("gpt-5") {
        return Some(ModelPricing {
            input_per_million: 2.5,
            cached_input_per_million: 0.25,
            cache_creation_per_million: 2.5,
            output_per_million: 15.0,
        });
    }
    if normalized.contains("claude-opus-4-7")
        || normalized.contains("claude-opus-4-6")
        || normalized.contains("claude-opus-4-5")
    {
        return Some(ModelPricing {
            input_per_million: 5.0,
            cached_input_per_million: 0.5,
            cache_creation_per_million: 6.25,
            output_per_million: 25.0,
        });
    }
    if normalized.contains("claude-opus-4") {
        return Some(ModelPricing {
            input_per_million: 15.0,
            cached_input_per_million: 1.5,
            cache_creation_per_million: 18.75,
            output_per_million: 75.0,
        });
    }
    if normalized.contains("claude-sonnet-4")
        || normalized.contains("claude-sonnet-3-7")
        || normalized.contains("claude-3-7-sonnet")
    {
        return Some(ModelPricing {
            input_per_million: 3.0,
            cached_input_per_million: 0.3,
            cache_creation_per_million: 3.75,
            output_per_million: 15.0,
        });
    }
    if normalized.contains("claude-haiku-4-5") {
        return Some(ModelPricing {
            input_per_million: 1.0,
            cached_input_per_million: 0.1,
            cache_creation_per_million: 1.25,
            output_per_million: 5.0,
        });
    }
    if normalized.contains("claude-haiku-3-5") {
        return Some(ModelPricing {
            input_per_million: 0.8,
            cached_input_per_million: 0.08,
            cache_creation_per_million: 1.0,
            output_per_million: 4.0,
        });
    }
    None
}

fn normalize_model_name(model_name: &str) -> String {
    model_name
        .trim()
        .trim_start_matches("codex/")
        .trim_start_matches("openai/")
        .trim_start_matches("claude/")
        .trim_start_matches("anthropic/")
        .to_ascii_lowercase()
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn charges_reasoning_tokens_at_output_rate() {
        let usage = TokenUsage {
            input_tokens: 1_000_000,
            cached_input_tokens: 500_000,
            cache_creation_tokens: 100_000,
            output_tokens: 10_000,
            reasoning_tokens: 5_000,
        };

        let cost = cost_from_pricing(
            ModelPricing {
                input_per_million: 2.0,
                cached_input_per_million: 0.2,
                cache_creation_per_million: 2.5,
                output_per_million: 10.0,
            },
            usage,
        );

        assert!((cost - 2.5).abs() < 0.000001);
    }

    #[test]
    fn leaves_unknown_models_unpriced() {
        let usage = TokenUsage {
            input_tokens: 1_000,
            cached_input_tokens: 0,
            cache_creation_tokens: 0,
            output_tokens: 1_000,
            reasoning_tokens: 0,
        };

        assert_eq!(estimate_cost_usd("mystery-provider/model", usage), None);
    }
}