Skip to content

gs2txt Documentation

LLM Providers

LLM Providers¶

gs2txt supports multiple LLM providers through a pluggable architecture.

Available Providers¶

Provider	Class	Models
OpenAI	`OpenAIProvider`	GPT-4, GPT-3.5-turbo
Anthropic	`AnthropicProvider`	Claude 3.5, Claude 3
LiteLLM	`LiteLLMProvider`	Any LiteLLM-supported model

OpenAIProvider¶

from gs2txt.llm import OpenAIProvider

provider = OpenAIProvider(
    api_key="sk-xxx",
    model_id="gpt-4",
    temperature=0.0,
    base_url=None  # Optional: custom endpoint
)

Parameters¶

Parameter	Type	Default	Description
`api_key`	`str`	Required	OpenAI API key
`model_id`	`str`	`"gpt-4"`	Model identifier
`temperature`	`float`	`0.0`	Sampling temperature
`base_url`	`str`	`None`	Custom API endpoint

Supported Models¶

gpt-4
gpt-4-turbo
gpt-4-turbo-preview
gpt-3.5-turbo

AnthropicProvider¶

from gs2txt.llm import AnthropicProvider

provider = AnthropicProvider(
    api_key="sk-ant-xxx",
    model_id="claude-sonnet-4-20250514",
    temperature=0.0
)

Parameters¶

Parameter	Type	Default	Description
`api_key`	`str`	Required	Anthropic API key
`model_id`	`str`	Required	Model identifier
`temperature`	`float`	`0.0`	Sampling temperature

Supported Models¶

claude-sonnet-4-20250514
claude-3-opus-20240229
claude-3-sonnet-20240229
claude-3-haiku-20240307

LiteLLMProvider¶

LiteLLM provides a unified interface for multiple LLM backends.

from gs2txt.llm import LiteLLMProvider

provider = LiteLLMProvider(
    api_key="your-api-key",
    model_id="gpt-4",
    temperature=0.0,
    base_url="https://your-litellm-server.com/"
)

Parameters¶

Parameter	Type	Default	Description
`api_key`	`str`	Required	API key
`model_id`	`str`	Required	Model identifier
`temperature`	`float`	`0.0`	Sampling temperature
`base_url`	`str`	`None`	LiteLLM server URL

Benefits¶

Unified API for multiple providers
Load balancing and fallbacks
Cost tracking
Rate limiting

Base Provider Interface¶

All providers implement BaseLLMProvider:

from gs2txt.llm.base import BaseLLMProvider

class BaseLLMProvider:
    def generate(self, messages: List[Dict]) -> str:
        """Generate response from messages."""
        raise NotImplementedError

    def validate_config(self) -> bool:
        """Validate provider configuration."""
        return True

Message Format¶

messages = [
    {"role": "system", "content": "You are a genomics expert..."},
    {"role": "user", "content": "Analyze these genes: TP53, MYC..."}
]

Creating Custom Providers¶

Implement your own provider:

from gs2txt.llm.base import BaseLLMProvider

class MyCustomProvider(BaseLLMProvider):
    def __init__(self, api_key: str, model_id: str, **kwargs):
        self.api_key = api_key
        self.model_id = model_id
        self.kwargs = kwargs

    def generate(self, messages: List[Dict]) -> str:
        # Your custom LLM call
        response = my_llm_call(
            messages=messages,
            api_key=self.api_key,
            model=self.model_id
        )
        return response.text

    def validate_config(self) -> bool:
        return bool(self.api_key and self.model_id)

Usage¶

provider = MyCustomProvider(
    api_key="...",
    model_id="my-model"
)
annotator = GeneSetAnnotator(llm_provider=provider)

Environment Variables¶

Variable	Provider	Description
`OPENAI_API_KEY`	OpenAI	API key
`ANTHROPIC_API_KEY`	Anthropic	API key
`LITELLM_API_KEY`	LiteLLM	API key

Error Handling¶

from gs2txt.llm import OpenAIProvider

provider = OpenAIProvider(api_key="invalid")

try:
    result = annotator.annotate(deg_df)
except Exception as e:
    print(f"LLM error: {e}")

Common errors:

Invalid API key
Rate limiting
Model not found
Network timeout