qai-sdk 0.1.21

Universal Rust SDK for AI Providers
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

//! # Multimodal Chat Example
//!
//! Demonstrates sending images and files alongside text to vision-capable models.

use qai_sdk::types::ImageSource;
use qai_sdk::*;

use qai_sdk::LanguageModel;
#[tokio::main]
async fn main() -> Result<()> {
    dotenvy::dotenv().ok();

    // ===================================================================
    // 1. Image URL input (OpenAI GPT-4o)
    // ===================================================================
    println!("=== Image URL Input (OpenAI) ===");
    let api_key = std::env::var("OPENAI_API_KEY").unwrap_or_default();
    let model = OpenAIModel::new(api_key);

    let prompt = Prompt {
        messages: vec![Message {
            role: Role::User,
            content: vec![
                Content::Text {
                    text: "What do you see in this image? Be brief.".to_string(),
                },
                Content::Image {
                    source: ImageSource::Url {
                        url: "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-702-702-702-702-702-702-702-702.jpg/1280px-Gfp-wisconsin-madison-the-nature-702.jpg".to_string(),
                    },
                },
            ],
        }],
    };

    let options = GenerateOptions {
        model_id: "gpt-4o-mini".to_string(),
        max_tokens: Some(200),
        temperature: None,
        top_p: None,
        stop_sequences: None,
        ..Default::default()
    };

    let result = model.generate(prompt, options).await?;
    println!("Vision response: {}\n", result.text);

    // ===================================================================
    // 2. Base64 image input (Anthropic Claude)
    // ===================================================================
    println!("=== Base64 Image Input (Anthropic) ===");
    let api_key = std::env::var("ANTHROPIC_API_KEY").unwrap_or_default();
    let model = AnthropicModel::new(api_key);

    // Example: 1x1 red PNG pixel as base64
    let tiny_png_base64 = "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8/5+hHgAHggJ/PchI7wAAAABJRU5ErkJggg==";

    let prompt = Prompt {
        messages: vec![Message {
            role: Role::User,
            content: vec![
                Content::Text {
                    text: "Describe this image.".to_string(),
                },
                Content::Image {
                    source: ImageSource::Base64 {
                        media_type: "image/png".to_string(),
                        data: tiny_png_base64.to_string(),
                    },
                },
            ],
        }],
    };

    let options = GenerateOptions {
        model_id: "claude-3-haiku-20240307".to_string(),
        max_tokens: Some(100),
        temperature: None,
        top_p: None,
        stop_sequences: None,
        ..Default::default()
    };

    let result = model.generate(prompt, options).await?;
    println!("Vision response: {}\n", result.text);

    // ===================================================================
    // 3. Multi-turn conversation with images (Google Gemini)
    // ===================================================================
    println!("=== Multi-turn with Image (Google) ===");
    let api_key = std::env::var("GOOGLE_GENERATIVE_AI_API_KEY").unwrap_or_default();
    let model = GoogleModel::new(api_key);

    let prompt = Prompt {
        messages: vec![
            Message {
                role: Role::User,
                content: vec![
                    Content::Text {
                        text: "Remember this image.".to_string(),
                    },
                    Content::Image {
                        source: ImageSource::Url {
                            url: "https://upload.wikimedia.org/wikipedia/commons/thumb/1/15/Cat_August_2010-4.jpg/1200px-Cat_August_2010-4.jpg".to_string(),
                        },
                    },
                ],
            },
            Message {
                role: Role::Assistant,
                content: vec![Content::Text {
                    text: "I see a cat in the image.".to_string(),
                }],
            },
            Message {
                role: Role::User,
                content: vec![Content::Text {
                    text: "What breed might it be?".to_string(),
                }],
            },
        ],
    };

    let options = GenerateOptions {
        model_id: "gemini-1.5-flash".to_string(),
        max_tokens: Some(200),
        temperature: None,
        top_p: None,
        stop_sequences: None,
        ..Default::default()
    };

    let result = model.generate(prompt, options).await?;
    println!("Multi-turn vision response: {}", result.text);

    Ok(())
}