openai-cost 0.1.0

Calculate OpenAI API call cost from a usage block. Cache-aware (cached_input_tokens), supports GPT-5, GPT-4.1, o3, o4 model families, BYO pricing override. No SDK dependency.
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

//! Per-model price table.
//!
//! All values are USD per 1,000,000 tokens, matching how OpenAI publishes
//! their rates. cached_input is typically 50% of fresh input.

use crate::normalize::normalize_model_id;
use crate::usage::Usage;

/// Per-model rates, USD per 1M tokens.
#[derive(Debug, Clone, Copy, PartialEq)]
pub struct Pricing {
    /// Fresh input tokens (no cache).
    pub input_per_mtok: f64,
    /// Output tokens.
    pub output_per_mtok: f64,
    /// Cached input tokens (cache hit).
    pub cached_input_per_mtok: f64,
}

impl Pricing {
    /// Compute USD cost for the given usage.
    pub fn cost_for(&self, usage: &Usage) -> f64 {
        (usage.input_tokens as f64 * self.input_per_mtok
            + usage.output_tokens as f64 * self.output_per_mtok
            + usage.cached_input_tokens as f64 * self.cached_input_per_mtok)
            / 1_000_000.0
    }
}

/// Built-in pricing table. Source: openai.com/api/pricing as of 2026-Q2.
/// VERIFY before billing.
///
/// Keys are normalized base aliases (no dated suffix). Use
/// [`default_pricing`] to look up by an aliased or dated model id.
pub const DEFAULT_PRICING_TABLE: &[(&str, Pricing)] = &[
    (
        "gpt-5",
        Pricing {
            input_per_mtok: 1.25,
            output_per_mtok: 10.0,
            cached_input_per_mtok: 0.125,
        },
    ),
    (
        "gpt-5-mini",
        Pricing {
            input_per_mtok: 0.25,
            output_per_mtok: 2.0,
            cached_input_per_mtok: 0.025,
        },
    ),
    (
        "gpt-5-nano",
        Pricing {
            input_per_mtok: 0.05,
            output_per_mtok: 0.4,
            cached_input_per_mtok: 0.005,
        },
    ),
    (
        "gpt-4.1",
        Pricing {
            input_per_mtok: 2.0,
            output_per_mtok: 8.0,
            cached_input_per_mtok: 0.5,
        },
    ),
    (
        "gpt-4.1-mini",
        Pricing {
            input_per_mtok: 0.4,
            output_per_mtok: 1.6,
            cached_input_per_mtok: 0.1,
        },
    ),
    (
        "gpt-4.1-nano",
        Pricing {
            input_per_mtok: 0.1,
            output_per_mtok: 0.4,
            cached_input_per_mtok: 0.025,
        },
    ),
    (
        "gpt-4o",
        Pricing {
            input_per_mtok: 2.5,
            output_per_mtok: 10.0,
            cached_input_per_mtok: 1.25,
        },
    ),
    (
        "gpt-4o-mini",
        Pricing {
            input_per_mtok: 0.15,
            output_per_mtok: 0.6,
            cached_input_per_mtok: 0.075,
        },
    ),
    (
        "o3",
        Pricing {
            input_per_mtok: 2.0,
            output_per_mtok: 8.0,
            cached_input_per_mtok: 0.5,
        },
    ),
    (
        "o3-mini",
        Pricing {
            input_per_mtok: 1.1,
            output_per_mtok: 4.4,
            cached_input_per_mtok: 0.55,
        },
    ),
    (
        "o4-mini",
        Pricing {
            input_per_mtok: 1.1,
            output_per_mtok: 4.4,
            cached_input_per_mtok: 0.275,
        },
    ),
];

/// Look up the price table entry for an OpenAI model id.
///
/// Accepts dated snapshot ids (`gpt-5-2026-04-01`) and resolves them back
/// to the base alias.
pub fn default_pricing(model_id: &str) -> Option<Pricing> {
    let key = normalize_model_id(model_id);
    DEFAULT_PRICING_TABLE
        .iter()
        .find(|(k, _)| *k == key)
        .map(|(_, p)| *p)
}