Utilities API Reference¶

Model Pricing¶

from flowgentra_ai import model_pricing

pricing = model_pricing("gpt-4")  # (input_price, output_price) per million tokens
# Returns None if model not found

Text Splitters¶

Specialized text splitters for different content types:

from flowgentra_ai import (
    RecursiveCharacterTextSplitter,
    MarkdownTextSplitter,
    HTMLTextSplitter,
    TokenTextSplitter,
    CodeTextSplitter,
)

Prompt & Parsers¶

from flowgentra_ai import PromptTemplate, JsonOutputParser, ListOutputParser

PromptTemplate¶

Template string formatting for prompts.

JsonOutputParser¶

Parse LLM output as JSON.

ListOutputParser¶

Parse LLM output as a list.

Routing Conditions¶

Build conditional routing logic declaratively:

from flowgentra_ai import ComparisonOp, Condition, ConditionBuilder

Evaluation¶

from flowgentra_ai import (
    EvaluationResult,
    EvaluationConfig,
    ScoringConfig,
    GradingConfig,
    EvaluationReport,
    NodeResult,
    evaluate_output_score,
)

Rerankers¶

Rerank search results for better relevance:

from flowgentra_ai import NoopReranker, RRFReranker, CrossEncoderReranker, LLMReranker

Class	Description
`NoopReranker`	Pass-through (no reranking)
`RRFReranker`	Reciprocal Rank Fusion
`CrossEncoderReranker`	Cross-encoder neural reranking
`LLMReranker`	LLM-based reranking

Advanced Nodes¶

Node types for complex graph patterns:

from flowgentra_ai import (
    BranchConfig,
    LoopNodeConfig,
    ParallelNodeConfig,
    SubgraphNodeConfig,
    JoinNodeConfig,
    JoinType,
    MergeStrategy,
)

MCP (Model Context Protocol)¶

from flowgentra_ai import MCPConfig

Visualization¶

from flowgentra_ai import (
    VisualizationConfig,
    visualize_graph,
    graph_to_dot,
    graph_to_mermaid,
)

RAG Configuration¶

from flowgentra_ai import (
    VectorStoreType,
    RAGConfig,
    VectorStoreConfig,
    EmbeddingsConfig,
    RetrievalSettings,
    PdfSettings,
    RAGGraphConfig,
)