Quality Settings

Learn how to configure transcription quality, understand the differences between each level, discover when to use advanced features, and optimize the balance between accuracy and cost.

Available Quality Levels

Basic Quality

Accuracy: 85-92% | Speed: 2-5x faster | Cost: Standard cost

Ideal for clear audio and general use.

Ideal for:

• Audio with good quality
• Single speaker
• Little background noise
• Personal or casual use
• Limited budget

Not recommended for:

• Audio with heavy noise
• Multiple overlapping speakers
• Complex technical jargon
• Critical professional use

Technical Specifications:

• Processing: Optimized base model
• Language model: Standard
• Noise reduction: Basic
• Context: 30 seconds

Advanced Quality

Accuracy: 92-96% | Speed: Standard speed | Cost: +50% of cost

Perfect balance between quality and cost.

Ideal for:

• Work meetings
• Professional interviews
• Content for publication
• Audio with average quality
• Professional use

Not recommended for:

• Very tight budget
• Extremely poor audio
• Non-critical transcriptions
• Casual use only

Technical Specifications:

• Processing: Advanced model with refinement
• Language model: Contextual + technical
• Noise reduction: Intelligent
• Context: 60 seconds

Premium Quality

Accuracy: 96-99% | Speed: 2-3x slower | Cost: +120% of cost

Maximum accuracy for critical professional use.

Ideal for:

• Medical/legal transcriptions
• Audio with heavy overlap
• Complex technical content
• Academic publications
• Compliance and auditing

Not recommended for:

• Casual use
• Limited budget
• Urgent results
• Simple audio

Technical Specifications:

• Processing: Premium multi-pass model
• Language model: Specialized + technical domains
• Noise reduction: Advanced AI
• Context: 120 seconds

Advanced Settings

Speaker Identification

Separates speech from different people.

| Option | Best for | |---|---| | Disabled | Single person or not important | | Enabled | Multiple people, meetings |

Timestamps

Adds time markers.

| Option | Best for | |---|---| | No timestamps | Simple running text | | Per sentence | Subtitles, synchronization | | Per word | Precise editing, analysis |

Profanity Filter

Removes or censors profanity.

| Option | Best for | |---|---| | Disabled | Faithful transcription | | Censor | Public content | | Remove | Corporate environment |

Smart Formatting

Improves punctuation and formatting.

| Option | Best for | |---|---| | Basic | Casual use | | Advanced | Publication, formality |

Next Steps

• Speaker Identification - How to separate voices in transcription
• Supported Formats - List of accepted formats
• Large Files - Tips for processing long files