Soup CLI vs Willow on IOS: Features, Pricing & Which Is Better (2026)
A side-by-side comparison of Soup CLI and Willow on IOS — features, pricing, and ideal use cases — to help you decide which AI tool fits your workflow.
S
Soup CLI
MePlay, Inc.
Open-source CLI that runs the whole LLM post-training stack — SFT, DPO, ORPO — on a 4GB laptop GPU.
Key features
- Whole Post-Training Stack: SFT, DPO, ORPO, SimPO, KTO, and more in one CLI.
- Low-VRAM Streaming: Fine-tune Llama-3.1-8B on a 4 GB GPU by streaming the base from RAM/NVMe.
- Auto-Configured Runs: Task, LR, epochs, and quantization derived from rules instead of grid search.
- Self-Healing Training: Detects and self-corrects reward hacking mid-run.
- One-Command Migration: `soup migrate` converts LLaMA-Factory, Axolotl, and Unsloth configs.
- Ship Gate: Every checkpoint is evaluated and either passes or is rejected before saving.
- Broad Ecosystem: Integrates with HuggingFace, Ollama, vLLM, DeepSpeed, Unsloth, ONNX, TensorRT, W&B.
- MLX + Apple Adapter: First-class Apple silicon support.
Best for
- Fine-tuning open-source LLMs on a consumer laptop GPU
- Post-training alignment (DPO/ORPO) without a rented A100
- Migrating existing LLaMA-Factory / Axolotl pipelines to a simpler workflow
- Producing evaluated, ship-gated checkpoints for internal deployment
- Researchers experimenting with 23 training methods without rewriting scripts
Willow on IOS
Willow Voice
Fast, context-aware speech-to-text dictation for Mac and iPhone with custom dictionaries and privacy-focused handling.
Key features
- Real-time Dictation: Converts spoken input into text on macOS and iPhone with immediate transcription for emails, documents, notes, and messages to speed up writing workflows.
- Context-Aware Processing: Uses contextual language understanding to improve accuracy and punctuation, adapting transcription to sentence structure and conversational context.
- Custom Dictionaries: Allows users to add domain-specific vocabulary, names, and technical terms so transcriptions reflect industry- or user-specific language correctly.
- Automatic Editing & Formatting: Applies automatic edits, punctuation, and formatting rules to raw transcribed text to reduce manual cleanup after dictation.
- App Integrations: Designed to work across common workflows and apps (email, documents, note-taking, messaging, and the Cursor editor) to insert transcribed text where users work.
- Privacy-Focused Handling: Emphasizes secure and private handling of voice data and transcription results to protect user information and sensitive content.
- Real-time speech-to-text dictation on iPhone and Mac
- Context-aware automatic edits to improve transcript quality
- Custom dictionaries and terminology support
- Privacy-focused operation with options for local/self-hosted inference
- Integration-ready: supports use in email, documents, note-taking, messaging, and developer workflows
- Willow Inference Server for self-hosted STT, TTS, LLM, and WebRTC inference
Best for
- Writing emails hands-free: Dictate long or short emails on Mac or iPhone to compose messages faster without switching to a keyboard.
- Meeting and lecture notes: Capture spoken content during meetings or lectures and get edited, punctuated notes ready for review and sharing.
- Document drafting and editing: Rapidly create drafts of reports, articles, or documents via voice, with automatic formatting reducing post-edit effort.
- Messaging and quick replies: Compose rapid, accurate message responses in chat and SMS apps using voice input on iPhone.
- Technical and domain-specific transcription: Use custom dictionaries to accurately transcribe industry jargon, code-related terms, names, and acronyms for developer or specialist workflows (e.g., Cursor integration).
- Accessibility and hands-free computing: Provide an accessible input method for users with mobility or dexterity impairments who need reliable speech-to-text on macOS and iOS.
- Hands-free email and document composition on iPhone and Mac
- Faster note-taking and meeting transcription
- Voice-driven messaging and chat input
- Accessibility for users needing speech input
- Enterprise/local deployment using Willow Inference Server for private on-premise transcription and TTS
- Developer integration for embedding STT/TTS/LLM capabilities into apps or real-time WebRTC flows