# aymenfurter/gpt-4o-latency-comparison architecture

> Architecture diagram of the GitHub repository aymenfurter/gpt-4o-latency-comparison, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2025-04-28). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/aymenfurter/gpt-4o-latency-comparison
- Repository: https://github.com/aymenfurter/gpt-4o-latency-comparison

## Overview

##### Project type and purpose
- This is a Python-based benchmarking toolkit with a simple Gradio “frontend” (app.py) and a modular “backend” consisting of an orchestrator (benchmark_runner.py), model-wrapper adapters (in models/), and utility modules (in utils/).
- It measures and compares latencies of various GPT-4o audio implementations by driving Azure OpenAI and Azure Speech APIs, analyzing returned audio via FFmpeg/pydub, and presenting metrics and audio samples in a web UI.

##### Main components
a. Gradio Interface (app.py)
b. Benchmark Orchestrator (benchmark_runner.py)
c. Model Adapter Layer (models/*.py)
d. Utility Layer (utils/audio_utils.py, utils/metrics.py, utils/client.py, etc.)
e. Configuration Loader (config.py + .env)
f. External Services:
- Azure OpenAI endpoints (GPT-4o realtime, audio preview, transcribe, mini-models)
- Azure Speech Services (speech-to-text, text-to-speech)
g. Local Audio Analysis Tools (FFmpeg/pydub binary invoked by audio_utils)

##### Relationships & data flow
1. User interacts with the Gradio Interface → selects prompt, models, iterations.
2. Gradio calls the Benchmark Orchestrator → loads config, instantiates each requested Model Adapter.
3. For each model and iteration:
- Adapter invokes Utils.client to call the appropriate Azure API.
- Azure returns text/audio streams.
- Adapter writes interim audio to disk or buffer.
- Utils.audio_utils calls FFmpeg/pydub to measure “time to playback” and “audio duration.”
- Utils.metrics aggregates timing data.
4. Orchestrator collects metrics & audio snippets → sends structured results back to Gradio.
5. Gradio renders charts, tables, and embeds audio players.

##### Architectural patterns & design principles
- Adapter Pattern for each GPT-4o variant (models/*)
- Single Orchestrator coordinating isolated benchmarks (benchmark_runner.py)
- Separation of concerns: UI (app.py) vs. Orchestration vs. Model logic vs. Utility functions
- Config-based deployment and environment credentials via config.py + .env

##### Detailed breakdown
– config.py reads .env
– benchmark_runner.py loop
5. Utils layer (audio_utils, metrics, client, exceptions) sits between adapters and local/external services.
7. Local audio analysis: FFmpeg/pydub invoked by audio_utils → returns durations.

## Diagram (Mermaid)

```mermaid
flowchart TB
    subgraph "User"
        Browser["User Browser"]
    end
    Browser --> UI

    UI["Gradio Web UI<br/>Gradio (app.py)"]:::ui
    click UI "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/app.py"
    UI --> Orchestrator

    subgraph "Backend"
        Config["Config Loader<br/>config.py + .env"]:::config
        click Config "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/config.py"
        Orchestrator["Benchmark Orchestrator<br/>benchmark_runner.py"]:::orch
        click Orchestrator "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/benchmark_runner.py"
        Config --> Orchestrator

        subgraph "Model Adapters"
            BaseAdapter["Base Adapter<br/>models/base_model.py"]:::adapter
            click BaseAdapter "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/base_model.py"
            AzureSpeech["Azure Speech Adapter<br/>models/azure_speech_model.py"]:::adapter
            click AzureSpeech "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/azure_speech_model.py"
            GPT4OAudio["GPT-4o Audio Adapter<br/>models/gpt4o_audio_model.py"]:::adapter
            click GPT4OAudio "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/gpt4o_audio_model.py"
            GPT4ORealtime["GPT-4o Realtime Adapter<br/>models/gpt4o_realtime_model.py"]:::adapter
            click GPT4ORealtime "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/gpt4o_realtime_model.py"
            GPT4OWhisper["GPT-4o Whisper Adapter<br/>models/gpt4o_whisper_model.py"]:::adapter
            click GPT4OWhisper "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/gpt4o_whisper_model.py"
            MiniTranscribe["GPT-4o Mini TTS Transcribe<br/>models/gpt4o_minitts_transcribe_model.py"]:::adapter
            click MiniTranscribe "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/gpt4o_minitts_transcribe_model.py"
            MiniMini["GPT-4o Mini Transcribe Mini<br/>models/gpt4o_minitts_mini_transcribe_model.py"]:::adapter
            click MiniMini "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/gpt4o_minitts_mini_transcribe_model.py"
            MiniWhisper["GPT4_1 Mini Whisper<br/>models/gpt41_mini_whisper_model.py"]:::adapter
            click MiniWhisper "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/models/gpt41_mini_whisper_model.py"
        end

        subgraph "Utils Layer"
            Client["API Client<br/>utils/client.py"]:::utils
            click Client "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/utils/client.py"
            AudioUtils["Audio Utils<br/>utils/audio_utils.py"]:::utils
            click AudioUtils "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/utils/audio_utils.py"
            Metrics["Metrics<br/>utils/metrics.py"]:::utils
            click Metrics "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/utils/metrics.py"
            Exceptions["Exceptions<br/>utils/exceptions.py"]:::utils
            click Exceptions "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/utils/exceptions.py"
        end

        Orchestrator --> BaseAdapter
        Orchestrator --> AzureSpeech
        Orchestrator --> GPT4OAudio
        Orchestrator --> GPT4ORealtime
        Orchestrator --> GPT4OWhisper
        Orchestrator --> MiniTranscribe
        Orchestrator --> MiniMini
        Orchestrator --> MiniWhisper

        BaseAdapter --> Client
        AzureSpeech --> Client
        GPT4OAudio --> Client
        GPT4ORealtime --> Client
        GPT4OWhisper --> Client
        MiniTranscribe --> Client
        MiniMini --> Client
        MiniWhisper --> Client

        Client --> OpenAI
        Client --> Speech

        audio_analysis["FFmpeg / pydub"]:::local
        click audio_analysis "https://github.com/aymenfurter/gpt-4o-latency-comparison/blob/main/utils/audio_utils.py"
        AudioUtils --> audio_analysis
        audio_analysis --> AudioUtils

        AudioUtils --> Metrics
        Metrics --> Orchestrator
        Client --> Exceptions

    end

    subgraph "External Cloud Services"
        OpenAI["Azure OpenAI Endpoints"]:::external
        Speech["Azure Speech Services"]:::external
    end

    Orchestrator --> UI

    classDef ui fill:#cce5ff,stroke:#0366d6
    classDef orch fill:#d4edda,stroke:#28a745
    classDef adapter fill:#d4edda,stroke:#28a745
    classDef utils fill:#fff3cd,stroke:#ffc107
    classDef config fill:#e2e3e5,stroke:#6c757d
    classDef external fill:#f0f0f0,stroke:#888888,stroke-dasharray:5 5
    classDef local fill:#f8d7da,stroke:#dc3545
```
