LLM Providers¶

Medium Converter supports various LLM providers for content enhancement. This page details each provider's setup and configuration.

OpenAI¶

Installation¶

pip install medium-converter[openai]

Configuration¶

# Environment variable
export OPENAI_API_KEY=your_api_key
export OPENAI_MODEL=gpt-4  # Optional, defaults to gpt-3.5-turbo

In Python:

from medium_converter import convert_article
from medium_converter.llm.config import LLMConfig, LLMProvider

llm_config = LLMConfig(
    provider=LLMProvider.OPENAI,
    model="gpt-4",
    api_key="your_api_key",
    temperature=0.7,
    max_tokens=None  # Use model default
)

await convert_article(
    url="https://medium.com/example-article",
    enhance=True,
    llm_config=llm_config
)

Available Models¶

gpt-3.5-turbo (default)
gpt-3.5-turbo-16k
gpt-4
gpt-4-turbo
gpt-4-32k

Anthropic¶

Installation¶

pip install medium-converter[anthropic]

Configuration¶

# Environment variable
export ANTHROPIC_API_KEY=your_api_key
export ANTHROPIC_MODEL=claude-3-sonnet-20240229  # Optional

In Python:

from medium_converter import convert_article
from medium_converter.llm.config import LLMConfig, LLMProvider

llm_config = LLMConfig(
    provider=LLMProvider.ANTHROPIC,
    model="claude-3-opus-20240229",
    api_key="your_api_key",
    temperature=0.5
)

await convert_article(
    url="https://medium.com/example-article",
    enhance=True,
    llm_config=llm_config
)

Available Models¶

claude-3-haiku-20240307 (default)
claude-3-sonnet-20240229
claude-3-opus-20240229
claude-2.1
claude-2.0

Google¶

Installation¶

pip install medium-converter[google]

Configuration¶

# Environment variable
export GOOGLE_API_KEY=your_api_key
export GOOGLE_MODEL=gemini-pro  # Optional, this is the default

In Python:

from medium_converter import convert_article
from medium_converter.llm.config import LLMConfig, LLMProvider

llm_config = LLMConfig(
    provider=LLMProvider.GOOGLE,
    model="gemini-pro",
    api_key="your_api_key",
    temperature=0.4
)

await convert_article(
    url="https://medium.com/example-article",
    enhance=True,
    llm_config=llm_config
)

Available Models¶

gemini-pro (default)
gemini-pro-vision
gemini-1.5-pro
gemini-1.5-flash

Mistral AI¶

Installation¶

pip install medium-converter[mistral]

Configuration¶

# Environment variable
export MISTRAL_API_KEY=your_api_key
export MISTRAL_MODEL=mistral-medium  # Optional, this is the default

In Python:

from medium_converter import convert_article
from medium_converter.llm.config import LLMConfig, LLMProvider

llm_config = LLMConfig(
    provider=LLMProvider.MISTRAL,
    model="mistral-large-latest",
    api_key="your_api_key",
    temperature=0.7
)

await convert_article(
    url="https://medium.com/example-article",
    enhance=True,
    llm_config=llm_config
)

Available Models¶

mistral-medium (default)
mistral-small-latest
mistral-large-latest
open-mixtral-8x7b
open-mistral-7b

Using LiteLLM¶

Medium Converter uses LiteLLM under the hood, which provides a unified interface to multiple LLM providers. This allows you to use any provider supported by LiteLLM by specifying the appropriate model name.

Installation¶

pip install medium-converter[llm]

Configuration¶

In Python:

from medium_converter import convert_article
from medium_converter.llm.config import LLMConfig

# Custom provider via LiteLLM
llm_config = LLMConfig(
    model="replicate/llama-2-70b-chat:58d078176e02c219e11eb4da5a02a7830a283b14cf8f94537af893ccff5ee781",
    api_key="your_replicate_api_key",
    extra_params={
        "provider": "replicate"
    }
)

await convert_article(
    url="https://medium.com/example-article",
    enhance=True,
    llm_config=llm_config
)

Provider Comparison¶

Provider	Strengths	Limitations	Cost
OpenAI	High quality, wide adoption	Higher cost, proprietary	$0.50-$30/million tokens
Anthropic	Great for long content, accurate	Higher cost, proprietary	$3-$15/million tokens
Google	Competitive pricing, reliable	Limited model options	$0.25-$1.50/million tokens
Mistral	Open weights options, competitive	Newer, less tested	$1-$8/million tokens
Local LLMs	Privacy, no API costs	Requires hardware, lower quality	Free (compute costs)