lm15

lm15 is a small, typed, provider-neutral interface for foundation-model requests, responses, streams, tools, media parts, endpoint APIs, errors, and canonical JSON serialization. This repository is its Python reference implementation.

What lm15 is — and deliberately is not. lm15 is a low-level foundation library: one canonical representation, exact serde for it, and adapters that translate it to and from each provider's wire format — stdlib-only, with its own HTTP transport (websockets is the single optional extra, for live sessions). It is NOT an opinionated user-facing API: no magic call(), no automatic tool loops, no DSL. lm15 is meant to be the dependency for libraries that want to build their own take on the right way to talk to AI systems in Python — you bring the opinions, lm15 brings every provider.

The public API is the top-level package: from lm15 import AnthropicLM, Request, Message, ... (see lm15/__init__.py for the full curated surface). Transport plumbing stays under lm15.transports, live sessions under lm15.live, and the conformance shim under lm15.vet.

The code blocks below are documentation that runs: every output block is the real, captured output of the example above it.

Measured by benchmarks/suite/run.py on Python 3.13.3 — methodology and full results in benchmarks/BENCHMARKS.md:

python3 -m pip install lm15
# Optional extra for websocket live sessions:
python3 -m pip install 'lm15[live]'

Or from source, for development:

git clone https://github.com/lm15-dev/lm15-python && cd lm15-python
python3 -m pip install -e '.[live]'

lm15 has zero required dependencies — it is stdlib-only, including its HTTP transports.

import os

from lm15 import Config, Message, OpenAILM, Request

lm = OpenAILM(api_key=os.environ["OPENAI_API_KEY"])

response = lm.complete(
    Request(
        model="gpt-4.1-mini",
        system="You are terse.",
        messages=(Message.user("Say hello in three words."),),
        config=Config(max_tokens=50, temperature=0.2),
    )
)

print(response.text)
print(response.finish_reason)
print(response.usage.total_tokens)

Hello there, friend.
stop
27

The mental model is one straight line:

Message parts → Message → Request → ProviderLM → Response
                              │
                              └── stream() → StreamEvent → materialized Response

The exact same Request shape drives the three first-party adapters:

import os

from lm15 import AnthropicLM, GeminiLM, Message, Request

providers = [
    AnthropicLM(api_key=os.environ["ANTHROPIC_API_KEY"]),
    GeminiLM(api_key=os.environ["GEMINI_API_KEY"]),
]

for lm in providers:
    response = lm.complete(
        Request(
            model={
                "anthropic": "claude-sonnet-4-5",
                "gemini": "gemini-3-flash-preview",
            }[lm.provider],
            messages=(Message.user("Say hello."),),
        )
    )
    print(lm.provider, response.text)

anthropic Hello! How can I help you today?
gemini Hello! How can I help you today?

And the same shape reaches every OpenAI-compatible server through OpenAIChatLM, the Chat Completions dialect adapter. A compat preset name — "ollama", "groq", "openrouter", "vllm", "sglang", ... — bundles that server's wire-format quirks and its default base_url, so a local Ollama is one constructor argument away:

from lm15 import Config, Message, OpenAIChatLM, Request

lm = OpenAIChatLM(api_key="ollama", compat="ollama")  # base_url -> http://localhost:11434/v1

response = lm.complete(
    Request(
        model="qwen3.5:0.8b",
        messages=(Message.user("Say hello in five words or fewer."),),
        config=Config(max_tokens=80, extensions={"reasoning_effort": "none"}),
    )
)

print(response.text)

Hello there! I'm ready to help. What would you like me to discuss?

Swap compat="groq" (plus your Groq key) or compat="openrouter" and the same request hits those servers; pass an explicit base_url to point a preset anywhere. Server-specific knobs ride in Config.extensions and pass through verbatim.

stream() yields typed StreamEvent objects. Text arrives as StreamDeltaEvent(delta=TextDelta(...)), and the stream is normalized across providers: exactly one StreamEndEvent ends the stream, carrying finish_reason and usage (mapping rule MAP-3).

import os

from lm15 import Message, OpenAILM, Request, StreamDeltaEvent, TextDelta

lm = OpenAILM(api_key=os.environ["OPENAI_API_KEY"])
request = Request(
    model="gpt-4.1-mini",
    messages=(Message.user("Write one short sentence about Montreal."),),
)

for event in lm.stream(request):
    if isinstance(event, StreamDeltaEvent) and isinstance(event.delta, TextDelta):
        print(event.delta.text, end="", flush=True)

Montreal is a vibrant, multicultural city in Canada known for its rich history and festivals.

To consume a stream into a full Response:

from lm15 import materialize_response

response = materialize_response(lm.stream(request), request)
print(response.text)

Montreal is a vibrant, multicultural city in Canada known for its rich history and cuisine.

The materialized Response is identical in shape to one from complete() — same message, finish_reason, usage, and provider_data.

lm15 distinguishes function tools that your application executes from provider-native built-in tools like web search. Here is the complete function-tool round-trip — model asks, you run your function, you answer back:

import os

from lm15 import FunctionTool, Message, OpenAILM, Request

lm = OpenAILM(api_key=os.environ["OPENAI_API_KEY"])

def get_weather(city: str) -> str:
    return f"Sunny and 22°C in {city}."

weather_tool = FunctionTool(
    name="get_weather",
    description="Get the current weather for a city.",
    parameters={
        "type": "object",
        "properties": {"city": {"type": "string"}},
        "required": ["city"],
    },
)

messages = (Message.user("What is the weather in Montreal?"),)
request = Request(model="gpt-4.1-mini", messages=messages, tools=(weather_tool,))

response = lm.complete(request)
for call in response.tool_calls:
    print(call.name, call.input)

get_weather {'city': 'Montreal'}

Now run your function and hand the result back. The model's tool-call turn is response.message; your answer is Message.tool({call_id: result}):

call = response.tool_calls[0]
result = get_weather(**call.input)

messages = (*messages, response.message, Message.tool({call.id: result}))
final = lm.complete(Request(model="gpt-4.1-mini", messages=messages, tools=(weather_tool,)))
print(final.text)

The weather in Montreal is sunny with a temperature of 22°C. Would you like to know the forecast for the coming days or any other information?

lm15 will never run the loop for you — that's your layer. This is the whole loop.

Built-in tools are provider-executed; you just declare them and read the results (citations come back as typed parts):

from lm15 import BuiltinTool, Message, Request

response = lm.complete(
    Request(
        model="gpt-4.1-mini",
        messages=(Message.user("Where will the 2028 Summer Olympics be held? One sentence, cite a source."),),
        tools=(BuiltinTool("web_search"),),
    )
)

print(response.text)
for citation in response.citations:
    print(citation.title, citation.url)

The 2028 Summer Olympics are scheduled to be held in Los Angeles, California, United States, from July 14 to 30, 2028. ([britannica.com](https://www.britannica.com/event/Los-Angeles-2028-Summer-Olympic-Games?utm_source=openai))
Los Angeles 2028 Summer Olympic Games | Bidding, Host, Venues, Planning, Sports, Marketing, & Facts | Britannica https://www.britannica.com/event/Los-Angeles-2028-Summer-Olympic-Games?utm_source=openai

Every adapter has an async mirror — AsyncOpenAILM, AsyncAnthropicLM, AsyncGeminiLM, AsyncOpenAIChatLM, AsyncClaudeCodeLM, AsyncOpenAICodexLM — with the same constructor fields, the same canonical Request in, and the same Response/stream events out. await is the only difference: complete() is async def, and stream() is an async for-able iterator of the same events.

import asyncio

from lm15 import (
    AsyncOpenAIChatLM,
    Config,
    Message,
    Request,
    StreamDeltaEvent,
    TextDelta,
)

async def main() -> None:
    lm = AsyncOpenAIChatLM(api_key="ollama", compat="ollama")
    request = Request(
        model="qwen3.5:0.8b",
        messages=(Message.user("Name two colors."),),
        config=Config(max_tokens=80, extensions={"reasoning_effort": "none"}),
    )

    response = await lm.complete(request)
    print(response.text)

    async for event in lm.stream(request):
        if isinstance(event, StreamDeltaEvent) and isinstance(event.delta, TextDelta):
            print(event.delta.text, end="", flush=True)
    print()

asyncio.run(main())

Two examples of natural and artificial colors are **red** and **blue**.
Two common names for a color are **red** (or crimson) and **blue** (often called indigo, cobalt, or azure). Other examples include green, yellow, purple, and brown.

The non-chat endpoints (embeddings, files, batch, image, audio, live) are sync-only for now; the async classes raise UnsupportedFeatureError for them rather than pretending. Async endpoint mirrors are planned.

The ordinary provider adapters use API keys that callers pass explicitly: OpenAILM(api_key=...), AnthropicLM(api_key=...), and GeminiLM(api_key=...).

lm15 also has explicit local-developer subscription adapters for users who are already signed in to provider CLIs. These adapters do not read API-key environment variables. They read local OAuth credentials created by the CLI and send provider-specific OAuth headers.

Use ClaudeCodeLM.from_claude_code() when Claude Code is installed and logged in as the same OS user:

from lm15 import ClaudeCodeLM, Config, Message, Request

lm = ClaudeCodeLM.from_claude_code()

response = lm.complete(
    Request(
        model="claude-fable-5",
        messages=(Message.user("Say hello briefly."),),
        config=Config(max_tokens=128),
    )
)

print(response.text)

The default credential path is ~/.claude/.credentials.json. If the credential is missing or expired, run Claude Code and log in again (claude, then /login if prompted).

ClaudeCodeLM always prepends the Claude Code system prompt required by this OAuth route:

You are Claude Code, Anthropic's official CLI for Claude.

If Request.system is also provided, lm15 keeps both: the required Claude Code prompt comes first, then the caller's system instruction.

Fable 5 note: Fable may spend part of max_tokens on hidden thinking, so a too-small budget can return no visible text with finish_reason="length". Use Config(max_tokens=128) or higher for non-trivial prompts.

Use OpenAICodexLM.from_codex_cli() when Codex CLI is installed and signed in with ChatGPT:

from lm15 import Message, OpenAICodexLM, Request

lm = OpenAICodexLM.from_codex_cli()

response = lm.complete(
    Request(
        model="gpt-5.5",
        messages=(Message.user("Say hello briefly."),),
    )
)

print(response.text)

The default credential path is ~/.codex/auth.json. OpenAICodexLM reads the local ChatGPT OAuth access token and account id from that file, then calls the Codex subscription endpoint. The Codex subscription backend is streaming-first, so complete() internally streams and materializes a normal Response.

Current Codex route note: lm15 intentionally omits max-token fields here because the verified local Codex route accepts the request shape without them; set output limits in your application layer if you need a hard cap.

These subscription adapters are intended for local interactive development, not server or CI deployments. Treat the credential files as secrets; do not print or log their bearer tokens.

Multimodal input uses typed media parts (ImagePart, AudioPart, DocumentPart, ...):

import os

from lm15 import ImagePart, Message, OpenAILM, Request, TextPart

lm = OpenAILM(api_key=os.environ["OPENAI_API_KEY"])

request = Request(
    model="gpt-4.1-mini",
    messages=(
        Message.user([
            TextPart("Describe this image in a few words."),
            ImagePart(
                url="https://raw.githubusercontent.com/github/explore/main/topics/react/react.png",
                media_type="image/png",
                detail="low",
            ),
        ]),
    ),
)

print(lm.complete(request).text)

This image shows a blue atomic symbol, often used to represent an atom or atomic energy.

Non-chat endpoints have separate request/response types — EmbeddingRequest, ImageGenerationRequest, AudioGenerationRequest, FileUploadRequest, BatchRequest, LiveConfig:

from lm15 import EmbeddingRequest

embeddings = lm.embeddings(
    EmbeddingRequest(
        model="text-embedding-3-small",
        inputs=("hello", "world"),
    )
)
print(len(embeddings.vectors), len(embeddings.vectors[0]))

2 1536

The serde functions convert every public lm15 type to canonical JSON-compatible dicts and back, exactly — this is the wire format the conformance corpus pins:

from lm15 import Message, Request, request_from_dict, request_to_dict

request = Request(model="gpt-4.1-mini", messages=(Message.user("Hi"),))
wire = request_to_dict(request)
round_tripped = request_from_dict(wire)
round_tripped == request

True

Provider-specific HTTP/API errors are normalized into one lm15 error hierarchy, so callers handle AuthError, RateLimitError, ContextLengthError, ... identically across providers:

import os

from lm15 import AuthError, Message, OpenAILM, ProviderError, RateLimitError, Request

lm = OpenAILM(api_key="not a key")

try:
    lm.complete(Request(model="gpt-4.1-mini", messages=(Message.user("Hi"),)))
except AuthError as exc:
    print("Check API key:", exc.env_keys)
except RateLimitError as exc:
    print("Retry later:", exc.retry_after)
except ProviderError as exc:
    print(exc.provider, exc.provider_code, exc.status, exc.request_id)

Check API key: ('OPENAI_API_KEY',)

ModelRegistry.discover() hydrates optional, advisory model metadata (pricing, context windows, capability hints) from installed catalog packages via the lm15.model_catalogs entry-point group — the aimo catalog is one such package. Hydrated metadata never changes what an adapter sends: requests are byte-identical with or without it. See docs/model-hydration.md for the contract.

• docs/design-rationale.md — why config=Config(...) instead of kwargs, why there is no automatic tool loop, why request extensions and response provider_data are different names on purpose.
• docs/serde-rules.md — the canonical JSON omission and round-trip rules.
• docs/mapping-rules.md — the provider mapping invariants (MAP-1, MAP-2, MAP-3, ...).
• Behavior is pinned by a cross-language conformance corpus: the sibling lm15-contract repository is the spec; this package is the reference implementation, not the authority — the full story is in How lm15 is specified.

The full documentation site — getting started, guides, runnable cookbooks, API reference, the specification pages, benchmarks, and the roadmap — lives at lm15-dev.github.io/lm15-python.

Fixture and conformance workflows, the doc-drift checker, the provider adapter development guide, and the useful-commands cheat sheet live in CONTRIBUTING.md.