swarmauri_llm_openai 0.11.0.dev1

Provider-specific Swarmauri import package for OpenAI chat, reasoning, tool-calling, transcription, and text-to-speech workflows.

Swarmauri OpenAI LLM

swarmauri_llm_openai provides provider-specific Swarmauri imports for OpenAI chat, reasoning, tool-calling, audio transcription, and text-to-speech workflows. The package exposes OpenAIModel, OpenAIReasonModel, OpenAIToolModel, OpenAIAudio, and OpenAIAudioTTS as direct provider-package imports while preserving the shared runtime implementations in swarmauri_standard.

The repo currently tracks OpenAI chat model families including gpt-5.x, gpt-4.1, gpt-4o, gpt-4-turbo, gpt-oss, reasoning families including o1, o3, and o4-mini, audio transcription models including whisper-1 and gpt-4o-*transcribe, and text-to-speech models including tts-1, tts-1-hd, and gpt-4o-mini-tts.

Why Use This Package?

• Keep OpenAI-specific imports explicit in Swarmauri applications.
• Use separate adapters for standard chat, reasoning, tool-calling, transcription, and TTS workflows.
• Reuse Swarmauri Conversation, toolkit, and component semantics across the OpenAI surface area.
• Track current OpenAI model families through one provider package boundary.

FAQ

What does swarmauri_llm_openai install?

It installs OpenAIModel, OpenAIReasonModel, OpenAIToolModel, OpenAIAudio, and OpenAIAudioTTS.

Which OpenAI capabilities are wrapped today?

The package covers standard chat completions, reasoning-oriented chat models, tool-calling workflows, audio transcription, and text-to-speech generation.

What is the difference between OpenAIModel and OpenAIReasonModel?

OpenAIModel targets general chat model families such as gpt-5.x, gpt-4.1, and gpt-4o. OpenAIReasonModel targets the reasoning families tracked in the repo, including o1, o3, and o4-mini variants.

When should I use OpenAIToolModel?

Use it when the application needs OpenAI tool calling through a Swarmauri toolkit. It is the provider-specific surface for tool-backed agent workflows.

What do OpenAIAudio and OpenAIAudioTTS do?

OpenAIAudio wraps audio transcription workflows. OpenAIAudioTTS wraps text-to-speech generation workflows.

Where should I verify current models and pricing?

Use the OpenAI section in docs/LLM_PROVIDER_MODEL_PRICING_LINKS.md, which points to the current OpenAI models, function-calling, audio, TTS, and pricing documentation.

Features

• OpenAIModel for sync, async, streaming, and batch chat completion workflows.
• OpenAIReasonModel for reasoning-model workflows across the repo-tracked o* families.
• OpenAIToolModel for toolkit-backed tool-calling and agent execution flows.
• OpenAIAudio for speech-to-text workflows against the tracked OpenAI transcription models.
• OpenAIAudioTTS for text-to-speech workflows against the tracked OpenAI TTS models.
• Compatibility with Python 3.10, 3.11, 3.12, 3.13, and 3.14.

Installation

uv add swarmauri_llm_openai

pip install swarmauri_llm_openai

Usage

Set OPENAI_API_KEY in your environment before creating any of the OpenAI adapters.

Chat Completion

import os

from swarmauri_llm_openai import OpenAIModel
from swarmauri_standard.conversations.Conversation import Conversation
from swarmauri_standard.messages.HumanMessage import HumanMessage

conversation = Conversation()
conversation.add_message(HumanMessage(content="Summarize Swarmauri in two sentences."))

model = OpenAIModel(
    api_key=os.environ["OPENAI_API_KEY"],
    name="gpt-5.5",
)
result = model.predict(conversation=conversation)

print(result.get_last().content)

Reasoning Model

import os

from swarmauri_llm_openai import OpenAIReasonModel
from swarmauri_standard.conversations.Conversation import Conversation
from swarmauri_standard.messages.HumanMessage import HumanMessage

conversation = Conversation()
conversation.add_message(HumanMessage(content="Reason through a release rollback plan step by step."))

model = OpenAIReasonModel(
    api_key=os.environ["OPENAI_API_KEY"],
    name="o3",
)
result = model.predict(conversation=conversation)

print(result.get_last().content)

Tool Calling

import os

from swarmauri_llm_openai import OpenAIToolModel
from swarmauri_standard.conversations.Conversation import Conversation
from swarmauri_standard.messages.HumanMessage import HumanMessage
from swarmauri_standard.toolkits.Toolkit import Toolkit
from swarmauri_standard.tools.AdditionTool import AdditionTool

conversation = Conversation()
conversation.add_message(HumanMessage(content="What is 19 + 23? Use the tool."))

toolkit = Toolkit(tools={"add": AdditionTool()})

model = OpenAIToolModel(
    api_key=os.environ["OPENAI_API_KEY"],
    name="gpt-5.5",
)
result = model.predict(
    conversation=conversation,
    toolkit=toolkit,
    tool_choice="auto",
)

print(result.get_last().content)

Audio Transcription

import os

from swarmauri_llm_openai import OpenAIAudio

audio_model = OpenAIAudio(
    api_key=os.environ["OPENAI_API_KEY"],
    name="whisper-1",
)

text = audio_model.predict("tests/static/test.mp3")
print(text)

Text To Speech

import os

from swarmauri_llm_openai import OpenAIAudioTTS

tts_model = OpenAIAudioTTS(
    api_key=os.environ["OPENAI_API_KEY"],
    name="tts-1",
)

audio_bytes = tts_model.predict("Swarmauri turns provider APIs into composable components.")
print(len(audio_bytes))

Examples

• Use OpenAIModel for general-purpose chat completion workflows.
• Use OpenAIReasonModel when you want the reasoning-model family explicitly.
• Use OpenAIToolModel when an agent needs toolkit-backed tool invocation.
• Use OpenAIAudio and OpenAIAudioTTS when the workflow crosses text, speech, and audio generation boundaries.

Related Packages

• swarmauri_llm_mistral
• swarmauri_llm_groq
• swarmauri_llm_hyperbolic
• swarmauri_llm_deepinfra
• swarmauri_llm_perplexity
• swarmauri_llm_playht

Foundational Swarmauri Packages

• swarmauri
• swarmauri_core
• swarmauri_base
• swarmauri_standard

More Documentation

• LLM provider model and pricing links
• OpenAI models
• OpenAI function calling
• OpenAI text to speech

Best Practices

• Keep OPENAI_API_KEY in environment variables or a secret manager.
• Pick the adapter that matches the workflow: chat, reasoning, tool-calling, transcription, or TTS.
• Validate model availability against current OpenAI docs before hard-coding a dated model variant in production.
• Use reasoning models deliberately, since they should be chosen for workflows that actually benefit from deeper multi-step reasoning.

License

Apache-2.0