Minimal building blocks for talking to LLMs from Erlang, with zero third-party
dependencies (only OTP inets
ssl
).
| Module | Role |
|---|---|
llm |
|
| one-call chat client for OpenAI (Responses API) and Anthropic, with tool-use and multimodal support | |
json_util |
|
| dependency-free JSON encode/decode |
%% rebar.config
{deps, [
{erlangchain, {git, "https://github.com/abhavk/erlangchain.git", {tag, "0.1.0"}}}
]}.
Set OPENAI_API_KEY
and/or ANTHROPIC_API_KEY
in the environment (a .env
file in the working directory is loaded automatically if present).
%% Simple completion (defaults to openai/small):
{ok, #{content := Text}} = llm:chat([#{role => user, content => <<"hello">>}]),
%% Pick provider + size:
{ok, Resp} = llm:chat(openai, big, Messages),
%% Tool use — pass tool specs, get back tool_calls to run and feed back:
{ok, #{tool_calls := Calls}} = llm:chat(openai, big, Messages, Tools).
Messages
are maps like #{role => system|user|assistant, content => binary()}
,
plus #{role => tool_result, tool_use_id => Id, content => Bin}
to return tool
output. See the header of src/llm.erl
for the full message/response shapes.
<<"{\"a\":1}">> = json_util:encode(#{<<"a">> => 1}),
#{<<"a">> := 1} = json_util:decode(<<"{\"a\":1}">>).
MIT — see LICENSE.
source & further reading
github.com — original article