README.md

In chatLLM: A Flexible Interface for 'LLM' API Interactions

chatLLM

Overview

chatLLM is an R package providing a single, consistent interface to multiple “OpenAI‑compatible” chat APIs (OpenAI, Groq, Anthropic, DeepSeek, Alibaba DashScope, Gemini, Grok and GitHub Models).

Key features:

• 🔄 Uniform API across providers
• 🗣 Multi‑message context (system/user/assistant roles)
• 🔁 Retries & backoff with clear timeout handling
• 🔈 Verbose control (verbose = TRUE/FALSE)
• ⚙️ Discover models via list_models()
• 🏗 Factory interface for repeated calls
• 🌐 Custom endpoint override and advanced tuning

Installation

From CRAN:

install.packages("chatLLM")

Development version:

# install.packages("remotes")  # if needed
remotes::install_github("knowusuboaky/chatLLM")

Setup

Set your API keys or tokens once per session:

Sys.setenv(
  OPENAI_API_KEY     = "your-openai-key",
  GROQ_API_KEY       = "your-groq-key",
  ANTHROPIC_API_KEY  = "your-anthropic-key",
  DEEPSEEK_API_KEY   = "your-deepseek-key",
  DASHSCOPE_API_KEY  = "your-dashscope-key",
  GH_MODELS_TOKEN    = "your-github-models-token",
  GEMINI_API_KEY     = "your-gemini-key",
  XAI_API_KEY        = "your-grok-key"
)

Usage

1. Simple Prompt

response <- call_llm(
  prompt     = "Who is Messi?",
  provider   = "openai",
  max_tokens = 300
)
cat(response)

2. Multi‑Message Conversation

conv <- list(
  list(role    = "system",    content = "You are a helpful assistant."),
  list(role    = "user",      content = "Explain recursion in R.")
)
response <- call_llm(
  messages          = conv,
  provider          = "openai",
  max_tokens        = 200,
  presence_penalty  = 0.2,
  frequency_penalty = 0.1,
  top_p             = 0.95
)
cat(response)

3. Verbose Off

Suppress informational messages:

res <- call_llm(
  prompt      = "Tell me a joke",
  provider    = "openai",
  verbose     = FALSE
)
cat(res)

4. Factory Interface

Create a reusable LLM function:

# Build a “GitHub Models” engine with defaults baked in
GitHubLLM <- call_llm(
  provider    = "github",
  max_tokens  = 60,
  verbose     = FALSE
)

# Invoke it like a function:
story <- GitHubLLM("Tell me a short story about libraries.")
cat(story)

5. Discover Available Models

# All providers at once
all_models <- list_models("all")
names(all_models)

# Only OpenAI models
openai_models <- list_models("openai")
head(openai_models)

6. Call a Specific Model

Pick from the list and pass it to call_llm():

anthro_models <- list_models("anthropic")
cat(call_llm(
  prompt     = "Write a haiku about autumn.",
  provider   = "anthropic",
  model      = anthro_models[1],
  max_tokens = 60
))

Troubleshooting

• Timeouts: increase n_tries / backoff or supply a custom .post_func with higher timeout().
• Model Not Found: use list_models("<provider>") or consult provider docs.
• Auth Errors: verify your API key/token and environment variables.
• Network Issues: check VPN/proxy, firewall, or SSL certs.

Contributing & Support

Issues and PRs welcome at https://github.com/knowusuboaky/chatLLM

License

MIT © Kwadwo Daddy Nyame Owusu - Boakye

Acknowledgements

Inspired by RAGFlowChainR, powered by httr and the R community. Enjoy!

chatLLM documentation built on Aug. 21, 2025, 6:03 p.m.