Audio Transcription Pipeline with Multi-Provider AI Processing

Transcription with speaker diarization plus AI post-processing for technical term correction.

Usage

# 1. Setup (one time)
./scripts/install_packages_and_venv.sh
cp setup_env.sh.example setup_env.sh
nano setup_env.sh  # Add API keys

# 2. Process single file
source venv/bin/activate && source setup_env.sh
./scripts/process_single.sh audio.mp3 --transcribers whisperx --processors opus

# 3. Multiple combinations (2 transcribers × 2 processors = 4 outputs)
./scripts/process_single.sh audio.mp3 --transcribers whisperx,assemblyai --processors opus,gemini

# 4. Cloud WhisperX (paid but no local GPU required)
./scripts/process_single.sh audio.mp3 --transcribers whisperx-cloud --processors opus

# 5. Batch process all MP3s
./scripts/process_all.sh --transcribers assemblyai --processors opus

Transcription Services

All services include speaker diarization (identifying who said what).

Service	Model	Type	Cost/hour	Speed
WhisperX	large-v3	Local GPU	FREE	5-10 min
WhisperX-Cloud	large-v3	Cloud API	$2.88	2-3 min
AssemblyAI	Best	Cloud API	$1.08	3-4 min

AI Post-Processors

Processor	Model	Cloud Service	Context Limit	Cost (Input/Output)	Best For
opus	Claude Opus 4.5	Anthropic	150K	$15/$75 per MTok	PREMIUM QUALITY - Best reasoning, long context
gemini	Gemini 3.0 Pro	Google	128K	~$1.25 per MTok	TECHNICAL - Superior technical preservation
deepseek	DeepSeek-V3.2	DeepSeek	128K	$0.28/$0.42 per MTok	COST-EFFECTIVE - Great quality at lowest cost
chatgpt	ChatGPT 5.2	OpenAI	varies	(see OpenAI pricing)	BALANCED - Strong general editing + instruction following

Setup

Requirements:

• Minimum: 8GB RAM, 50GB disk (CPU-only, slow)
• Recommended: NVIDIA GPU 6GB+ VRAM, 16GB RAM, 50GB disk

API Keys in setup_env.sh:

export HF_TOKEN="hf_..."              # Required for WhisperX diarization
export ANTHROPIC_API_KEY="sk-ant-..." # For Claude Opus post-processing
export GOOGLE_API_KEY="..."           # For Gemini post-processing
export DEEPSEEK_API_KEY="..."         # For DeepSeek post-processing
export OPENAI_API_KEY="sk-..."         # For ChatGPT 5.2 post-processing
export ASSEMBLYAI_API_KEY="..."       # Optional: AssemblyAI transcription
export REPLICATE_API_TOKEN="..."      # Optional: Cloud WhisperX

Output Files

Naming: {basename}_{transcriber}_{processor}.{txt|md}

Example: interview_assemblyai_opus.txt

Transcribers: whisperx, whisperx-cloud, assemblyai

Processors: opus, gemini, deepseek, chatgpt

License

GPL-3.0 - See LICENSE file.