TTS availability and billing update, October 6, 2026
OpenAI's tts-1, tts-1-hd and dated gpt-4o-mini-tts snapshots (2025-03-20 and 2025-12-15) are deprecated, not yet discontinued. Shutdown is scheduled for January 6, 2027. Existing character-price estimates remain for migration budgeting; they are not recommendations for new integrations. OpenAI recommends gpt-realtime-2.1-mini, whose audio-token billing cannot be compared as a character rate. See the official migration notice.
Microsoft publishes MAI-Voice-2.1 at $22/M characters and MAI-Voice-2.1-Flash at $15/M characters, both with 23 languages. These are distinct from older Azure TTS tiers. Azure Speech labels MAI-Voice integration Public Preview; voice cloning is gated and requires consent. Microsoft's published rates are research-based here, not hands-on quality rankings.
Text-to-Speech API Cost Calculator
Text-to-Speech API Cost Ranking
Every provider ranked by cost for a typical workload: 500K chars/day, neural-quality voices.
Top Picks by Volume
Strategy: Quality-Volume Routing
Use quality-volume routing — route high-volume, low-stakes content to cheap standard voices, and premium content to neural/premium voices.
Quality-volume routing saves 50% compared to using premium voices for all content. Most output is internal or low-stakes — only customer-facing and marketing content needs premium voices.
Stop guessing — get exact Text-to-Speech API costs
No signup required to 128-model comparison, migration code snippets, PDF reports, price alerts, and cost monitoring. ✅ All tools free.
Free Tools →Find the cheapest model for your TTS workload
Enter your usage and see all providers ranked by cost. Free, no signup.
Open Savings Calculator →Key Factors When Choosing a TTS API
- Per-character pricing: TTS APIs charge per character of input text. One hour of speech ≈ 500K-700K characters (at ~150 words/minute). Standard voices are 3-4× cheaper than neural voices.
- Voice quality tiers: Standard ($4/1M chars) — robotic but functional. Neural ($16/1M chars) — natural-sounding, good for most uses. Premium ($15-30/1M chars) — human-like, best for customer-facing content.
- SSML support: If you need pronunciation control, pauses, or emphasis, check SSML support. OpenAI TTS and Google have the best SSML implementations.
- Voice cloning: ElevenLabs offers voice cloning on paid tiers. If you need a custom brand voice, this is the main option. Other providers offer pre-built voices only.
- Languages: Google and Amazon support 30+ languages. OpenAI TTS supports ~10 languages with high quality. ElevenLabs focuses on English and a few major languages.
- Streaming: For real-time applications (chatbots, assistants), streaming TTS reduces time-to-first-audio. OpenAI, Google, and ElevenLabs all support streaming.
Related Tools
- Free MCP Server — Query live pricing data in Claude Code, Cursor
- other AI tools
- Savings Calculator — See how much you can save by switching models
- Cost Explorer — See all 128 models ranked by your usage
- State of AI API Pricing 2026 — 128 models compared, 5 key trends, 40-96% savings strategies
- Cheapest AI API Finder — Find the absolute cheapest model
- Migration Checklist — 9 provider migration routes with code examples
- Deprecation Tracker — 6 deprecated models and migration paths
- Budget Planner — Describe your app, get instant cost estimates
Related Reading
- Cheapest LLM APIs in 2026 — Full ranking of every model
- Cheapest AI API for Speech-to-Text — STT cost comparison
- Cheapest AI API for Chatbots — Chatbot cost comparison