oxi-ai 0.13.0

Unified LLM API — multi-provider streaming interface for AI coding assistants
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

#![allow(unused_doc_comments)]
#![warn(missing_docs)]
#![warn(clippy::unwrap_used)]
#![allow(clippy::unwrap_used_in_tests)]

//! oxi-ai: Unified LLM API for oxi
//!
//! This crate provides a unified interface for interacting with multiple LLM providers.
//! It handles streaming, tool calling, context management, and cross-provider handoffs.

mod compaction;
mod context;
pub mod env_api_keys;
mod error;
mod high_level;
mod messages;
pub mod oauth;
pub mod provider_registry;
mod providers;

pub mod register_builtins { pub use crate::providers::register_builtins::*; }
pub mod secret;
mod tools;
mod transform;
mod types;
pub mod utils;

/// Standard imports for oxi-ai usage.
pub mod prelude {
    pub use crate::compaction::{
        CompactedContext, CompactionManager, CompactionStrategy, Compactor, LlmCompactor,
    };
    pub use crate::context::Context;
    pub use crate::error::{Error, Result};
    pub use crate::messages::*;
    pub use crate::providers::{Provider, ProviderEvent, StreamOptions};
    pub use crate::tools::{validate_args, Tool};
    pub use crate::types::*;
}

// Re-export main types

/// Provider-specific error type for LLM operations.
pub use crate::error::ProviderError;

/// Shared conversation context.
pub use context::Context;

/// Result type alias for oxi-ai operations.
pub use error::{Error, Result};

/// Message types for constructing conversations.
pub use messages::*;

/// Cache retention control for provider requests.
pub use providers::CacheRetention;

/// Provider trait, streaming options, and provider registry.
pub use providers::{get_provider, register_provider, unregister_provider, custom_provider_names, Provider, ProviderEvent, ProviderRegistry, StreamOptions};

/// Built-in provider helpers (re-exported from providers).
pub use providers::register_builtins::{
    create_builtin_provider, get_builtin_provider,
    get_provider_env_key, get_provider_env_keys, get_all_provider_names, is_builtin_provider,
};

/// OpenAI-compatible provider implementation.
pub use providers::OpenAiProvider;

/// Model fetching utilities (async and blocking).
pub use providers::model_fetch::{fetch_models_async, fetch_models_blocking};

/// OpenAI Responses API provider.
pub use providers::OpenAiResponsesProvider;

/// Tool definition and argument validation.
pub use tools::{progress_callback, validate_args, ProgressCallback, Tool, ToolValidationError};

/// Core type definitions (tokens, cost, etc.).
pub use types::*;

// High-level API

/// Token estimation and context usage helpers.
pub use high_level::tokens::{context_usage, estimate, estimate_words};

/// High-level completion and token estimation.
pub use high_level::{complete, estimate_tokens};

// Context compaction

/// Compaction strategies and managers for long conversations.
pub use compaction::{
    CompactedContext, CompactionManager, CompactionStrategy, Compactor, LlmCompactor,
};

// Cross-provider message transformation

/// Message transformation between provider formats.
pub use transform::{
    anthropic_to_google, anthropic_to_openai, google_to_openai, normalize_tool_call_id,
    openai_to_anthropic, transform_messages, transform_messages_for_model, TransformOptions,
};

// Model registry (runtime mutable registry)
mod model_registry;

/// Runtime model registry for dynamically registered models.
///
/// Unlike the static `model_db`, this supports adding/removing models at runtime.
pub use model_registry::{get_model, get_models, get_providers, register_model, unregister_model, lookup_model, ModelRegistry};

// Static model database (comprehensive)
pub mod model_db;

/// Static database of known models with cost and modality info.
///
/// Provides comprehensive model listings, filtering, and search capabilities.
pub use model_db::{
    get_all_models, get_cheapest_models, get_model_entry, get_provider_models,
    get_reasoning_models, get_vision_models, model_count, search_models, ModelEntry,
};

/// Re-export AssistantMessage from messages
pub use messages::AssistantMessage;

// Environment-based API key resolution

/// Utilities for discovering API keys from the environment.
pub use env_api_keys::{find_env_keys, get_all_env_keys, get_env_api_key};

// Provider authentication registry

/// OAuth token and API key management for providers.
pub use provider_registry::{OAuthTokenInfo, ProviderAuth, ProviderAuthRegistry};