Speech-to-Text AI tools: compare listings

All Speech-to-Text tools (13)

All Tags

A2A Protocol Agent Framework Agent Orchestration Agent Runtime Agent Wallet AI Agent AI Assistant AI Content AI Payments AI Search Alibaba Anthropic API Automation AWS BAAI Black Forest Labs Browser Automation ByteDance Chatbot Claude Claude Code CLI Cloud Cloudflare Code Assistant Code Quality Code Review Codex Coding Agent Cohere Communication Computer Use Crypto & Web3 Cursor Data Science Data Visualization Database Debugging Deep Research Deepgram DeepSeek Deployment Design Developer Tools Document Processing Documentation Education ElevenLabs Embedding Enterprise Evaluation Fine-Tuning FLUX Frontier Model Gemini Gemma Git Go Google GPT GPU Grok Health Hugging Face IBM Image Generation Integration Jina AI Kimi Knowledge Base LangChain LangGraph Linux Llama Local AI Long Context macOS Markdown MCP Meta Microsoft MiniMax Mistral Mixedbread AI Mixture of Experts Model Routing Moonshot AI Multi-Agent Multilingual Multimodal No-Code / Low-Code Nous Research NVIDIA Observability Obsidian Open Source Open Weights OpenAI OpenClaw OpenCode OpenHands OpenRouter Perplexity Presentations Privacy Productivity Project Management Protocol Python Qwen RAG Real-Time Reasoning Replit Reranking Research Rust Sandbox Security Self-Hosted Social Media Speech-to-Text

Items tagged with Speech-to-Text (13)

Mistral Voxtral Mini Transcribe 2

Models

Mistral Voxtral Mini Transcribe 2 (voxtral-mini-2602) is a transcription-only audio model released February 4, 2026, priced at $0.003 per minute.

#Mistral

800000DR: 90AS: 88

Mistral Voxtral Mini Transcribe Realtime

Models

Voxtral Mini Transcribe Realtime is Mistral's open-weight live speech-to-text model (Apache 2.0, 13 languages), released February 4, 2026 at $0.006/min.

#Mistral

800000DR: 90AS: 88

Whisper V3

Models

Whisper large-v3 is OpenAI's open-weight speech recognition and translation model, with 99 languages, 128 Mel bins and 10 to 20% fewer errors than large-v2.

#OpenAI

8000DR: 95AS: 92

Deepgram Nova-2

Models

Deepgram Nova-2 is a previous speech-to-text model, still listed for languages and filler-word identification that Nova-3 does not yet cover.

#Deepgram

2500DR: 85AS: 83

BaoCut

Generative AI

BaoCut combines local transcription, bilingual subtitles, translated dubbing, and agent-assisted video editing on Apple Silicon Macs.

#Video

—DR: —AS: —

Index-Translate

Models

Bilibili's open multilingual translation family: 150-language text, speech subtitles, controlled dubbing, and long-document translation built on Qwen3.5.

#Multilingual

—DR: —AS: —