SlimX is a slim, intuitive, lightweight Python library for calling LLMs and building LLM systems.

It is intentionally designed around two clearly separated APIs:

• High-level API (slimx) — “1‑minute productivity”: llm(...), .stream(...), .json(...), tools, retries.
• Low-level API (slimx.low) — “systems builder primitives”: explicit Client, ChatRequest, Message, provider registry, middleware.

SlimX also supports multiple providers (OpenAI, Anthropic, Ollama) and provider plugins (3rd-party providers without modifying core).

On Debian/Ubuntu you may hit externally-managed-environment (PEP 668) if you try to use system pip. Use uv, which manages an isolated environment cleanly.

git clone https://github.com/slimx-ai/slimx.git
cd slimx
uv sync --all-extras

unzip slimx_v0_4.zip
cd slimx_v0_4
uv sync --all-extras
uv run python examples/quickstart_openai.py

uv sync reads pyproject.toml and (optionally) uv.lock. If uv.lock is present and committed, installs are reproducible.

export OPENAI_API_KEY="..."
# optional:
export OPENAI_BASE_URL="https://api.openai.com/v1"

export ANTHROPIC_API_KEY="..."
# optional:
export ANTHROPIC_BASE_URL="https://api.anthropic.com"
export ANTHROPIC_VERSION="2023-06-01"

export OLLAMA_BASE_URL="http://localhost:11434"

from slimx import llm
m = llm("openai:gpt-4.1-nano", temperature=0.2)
res = m("Write a haiku about fog and streetlights.")
print(res.text)

Streaming:

for ev in m.stream("Tell a short story in 5 lines."):
    if ev.type == "token":
        print(ev.text, end="", flush=True)
print()

Tools (auto-loop):

from slimx import llm, tool

@tool
def add(a: int, b: int) -> int:
    "Add two integers."
    return a + b

m = llm("openai:gpt-4.1-nano", tools=[add], tool_runtime="auto")
print(m("What is 12 + 30?").text)

Structured output:

from dataclasses import dataclass
from slimx import llm

@dataclass
class City:
    name: str
    country: str

m = llm("openai:gpt-4.1-nano")
res = m.json("Paris is in France.", schema=City)
print(res.data)

This bundle includes GitHub Actions:

• CI (.github/workflows/ci.yml)
• Docs deploy to GitHub Pages (docs.yml)
• PyPI publish on tag (publish.yml)

See README.md and docs/ for details.