unigateway 1.2.1

Lightweight, local-first LLM gateway for developers. A stable, single-binary unified entry point for all your AI tools and models.
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

use std::sync::Arc;

use axum::{http::HeaderMap, response::Response};
use serde_json::Value;

use crate::types::AppState;

mod execution_flow;
mod request_flow;
mod response_flow;

use self::execution_flow::{
    execute_prepared_anthropic_chat, execute_prepared_openai_chat,
    execute_prepared_openai_embeddings, execute_prepared_openai_responses,
};
use self::request_flow::{prepare_and_parse_anthropic_request, prepare_and_parse_openai_request};

pub(super) async fn handle_openai_chat_request(
    state: &Arc<AppState>,
    headers: &HeaderMap,
    payload: &Value,
) -> Response {
    let (prepared, request) = match prepare_and_parse_openai_request(
        state,
        headers,
        payload,
        unigateway_runtime::protocol::openai_payload_to_chat_request,
    )
    .await
    {
        Ok(parts) => parts,
        Err(resp) => return resp,
    };

    execute_prepared_openai_chat(state, &prepared, &request).await
}

pub(super) async fn handle_openai_responses_request(
    state: &Arc<AppState>,
    headers: &HeaderMap,
    payload: &Value,
) -> Response {
    let (prepared, request) = match prepare_and_parse_openai_request(
        state,
        headers,
        payload,
        unigateway_runtime::protocol::openai_payload_to_responses_request,
    )
    .await
    {
        Ok(parts) => parts,
        Err(resp) => return resp,
    };

    execute_prepared_openai_responses(state, &prepared, request).await
}

pub(super) async fn handle_anthropic_messages_request(
    state: &Arc<AppState>,
    headers: &HeaderMap,
    payload: &Value,
) -> Response {
    let (prepared, request) = match prepare_and_parse_anthropic_request(
        state,
        headers,
        payload,
        unigateway_runtime::protocol::anthropic_payload_to_chat_request,
    )
    .await
    {
        Ok(parts) => parts,
        Err(resp) => return resp,
    };

    execute_prepared_anthropic_chat(state, &prepared, &request).await
}

pub(super) async fn handle_openai_embeddings_request(
    state: &Arc<AppState>,
    headers: &HeaderMap,
    payload: &Value,
) -> Response {
    let (prepared, request) = match prepare_and_parse_openai_request(
        state,
        headers,
        payload,
        unigateway_runtime::protocol::openai_payload_to_embed_request,
    )
    .await
    {
        Ok(parts) => parts,
        Err(resp) => return resp,
    };

    execute_prepared_openai_embeddings(state, &prepared, &request).await
}