Cartesia (Sonic)

Best for: Prototypes and side projects - free to start, no sales call; Regulated or enterprise workloads - compliance attestations and an enterprise plan; AI agents and automation - an agent-ready surface (MCP / llms.txt)

Pricing model

Hybrid (base + usage) [2]cartesia.ai/pricing“Free $0/month 20K credits/month; Pro $5/month 100K credits/month; Startup $49/month 1.25M credits/month; Scale $299/month 8M credits/month”

Published pricing

✓ Yes

Free tier

✓ Yes [3]cartesia.ai/pricing“Free - $0/month; 20K credits/month; Text to Speech & Speech to Text included; Instant voice cloning included”

Free tier details

Free plan at $0/month includes 20,000 credits/month (approximately 20,000 characters of TTS at 1 credit per character). Includes instant voice cloning. Does not include professional voice cloning or commercial use license. [4]cartesia.ai/pricing“Free & Pro: Instant voice cloning available. Startup & Scale: Professional voice cloning added.”

Self-serve signup

✓ Yes

Requires sales call

✗ No

Enterprise plan

✓ Yes [5]cartesia.ai/pricing“Enterprise - Custom pricing; Custom credits & agent usage with volume pricing; Includes: Custom concurrency, DPAs/BAAs, Shared Slack, SSO, Security questionnaires”

• Real-time streaming
• Voice cloning
• Voice design
• Multilingual voices
• Word timestamps

Supported actions

synthesize_speech, streaming_tts, websocket_tts, instant_voice_cloning, professional_voice_cloning, word_timestamps, phoneme_timestamps, speech_to_speech, multilingual_synthesis, emotion_control, speed_control, volume_control, pronunciation_dictionaries, voice_design, voice_localization, voice_infill, voice_changer [6]docs.cartesia.ai/api-reference/tts/websocket“Word-level and phoneme-level timestamps; Flush acknowledgments; Completion signals”docs.cartesia.ai/api-reference/tts/sse“timestamps, phoneme_timestamps, pronunciation dictionary, generation config (volume, speed, emotion), context_id”

Regions

US, EU, global (cloud), on-premises, on-device

Languages

American English, British English, Australian English, French (France), French (Canada), German, Spanish (Castilian), Spanish (Mexican), Italian, Portuguese (Brazilian), Russian, Japanese, Korean, Mandarin Chinese, Dutch, Hindi, Bengali, Bulgarian, Croatian, Czech, Danish, Finnish, Georgian, Greek, Gujarati, Hebrew, Hungarian, Indonesian, Kannada, Malayalam, Marathi, Norwegian, Polish, Punjabi, Romanian, Slovak, Swedish, Tagalog, Tamil, Telugu, Thai, Turkish, Ukrainian, Vietnamese, Emirati Arabic, Malay, Arabic [7]docs.cartesia.ai/build-with-cartesia/tts-models/latest“Ranked #1 for naturalness, with sub-90ms latency and native support for 42 languages”cartesia.ai/languages“47 languages and regional variants: American English, British English, Australian English, French (France), French (Canada), German, Castilian Spanish, Mexican Spanish, Italian, Brazilian Portuguese, Russian, Japanese, Korean, Mandarin Chinese, Dutch, Hindi, Bengali, Bulgarian, Croatian, Czech, Danish, Finnish, Georgian, Greek, Gujarati, Hebrew, Hungarian, Indonesian, Kannada, Malayalam, Marathi, Norwegian, Polish, Punjabi, Romanian, Slovak, Swedish, Tagalog, Tamil, Telugu, Thai, Turkish, Ukrainian, Vietnamese, Emirati Arabic, Malay, Arabic”

Input types

plain text

Output types

pcm_f32le, pcm_s16le, pcm_mulaw, pcm_alaw, wav, mp3, streaming chunks (SSE), streaming chunks (WebSocket) [8]docs.cartesia.ai/api-reference/tts/sse“Encoding options: pcm_f32le, pcm_s16le, pcm_mulaw, pcm_alaw; Sample rates: 8000, 16000, 22050, 24000, 44100, 48000 Hz”docs.cartesia.ai/api-reference/tts/bytes“RAW: PCM with custom encoding/sample rate; WAV: PCM with header wrapper; MP3: compressed with bitrate selection”

Webhooks

✗ No [9]docs.cartesia.ai/line/infrastructure/observability“Cartesia sends five types of webhook events throughout the call lifecycle: call_started, call_completed, call_failed, call_turn, post_call_analysis”

Sandbox / test mode

✗ No

SDK languages

Python, JavaScript/TypeScript [10]docs.cartesia.ai/tools/client-libraries“JavaScript/TypeScript: @cartesia/cartesia-js (https://github.com/cartesia-ai/cartesia-js). Python: cartesia (https://github.com/cartesia-ai/cartesia-python)”

MCP server

✓ Yes [11]docs.cartesia.ai/integrations/mcp“Run cartesia-mcp with uvx to use Cartesia TTS, STT, voices, and pronunciation dictionaries from MCP clients like Cursor and Claude Code.”github.com/cartesia-ai/cartesia-mcp“The official Cartesia MCP Server — exposes Cartesia APIs over the Model Context Protocol (MCP) so clients such as Cursor, Claude Desktop, and OpenAI Agents can list voices, run TTS, clone voices, infill audio, and more.”

SOC 2

SOC 2 Type II [12]cartesia.ai/blog/gdpr-compliance“SOC 2 Type II, PCI-DSS (service level provider), HIPAA, and optional Zero Data Retention”cartesia.ai/sonic“SOC 2 Type 2 certified”

HIPAA

✓ Yes [13]cartesia.ai/blog/gdpr-compliance“SOC 2 Type II, PCI-DSS (service level provider), HIPAA, and optional Zero Data Retention”cartesia.ai/sonic“HIPAA compliant”

GDPR

✓ Yes [14]cartesia.ai/legal/dpa“The DPA explicitly incorporates GDPR requirements, defining 'European Data Protection Laws' as 'the General Data Protection Regulation (EU) 2016/679'”cartesia.ai/blog/gdpr-compliance“our text-to-speech platform is now GDPR compliant”

ISO 27001

– Unknown [15]cartesia.ai/sonic“SOC 2 Type 2 certified, HIPAA compliant, PCI certified, GDPR”

PCI DSS

✓ Yes [16]cartesia.ai/blog/gdpr-compliance“SOC 2 Type II, PCI-DSS (service level provider), HIPAA, and optional Zero Data Retention”cartesia.ai/sonic“PCI certified”

Published SLA

✗ No [17]cartesia.ai/vs/cartesia-vs-elevenlabs“Enterprise-grade reliability with 99.9% uptime, SOC2 compliance, and full on-premises support.”

Rate limits

TTS concurrent requests: Free=2, Pro=3, Startup=5, Scale=15, Enterprise=custom. WebSocket connections limited to 10x concurrency limit. Idle WebSocket connections closed after 5 minutes. Exceeding limits returns HTTP 429. First audio byte latency: 90ms (Sonic 3.5). [18]docs.cartesia.ai/use-the-api/concurrency-limits-and-timeouts“We limit the number of parallel TTS WebSocket connections to 10X your concurrency limit.”docs.cartesia.ai/use-the-api/concurrency-limits-and-timeouts“Free: 2 TTS concurrent connections; Pro: 3 TTS; Startup: 5 TTS; Scale: 15 TTS; Enterprise: Custom”

Known restrictions

Voice cloning requires explicit consent from the voice owner per Acceptable Use Policy, Cannot post audio of deceased persons, political candidates, or others without express permission, No SSML support documented, Professional voice cloning available on Startup plan and above only; instant voice cloning available on all plans including Free, Free plan does not include commercial use license, Infinite request length supported (no per-request character cap documented), Voice sample recordings may be retained by Cartesia solely for voice cloning functionality [19]cartesia.ai/pricing“Free & Pro: Instant voice cloning available. Startup & Scale: Professional voice cloning added.”cartesia.ai/legal/acceptable-use.html“You may only submit your own voice and audio recordings or those of others with explicit consent”

• Docs
• Status page

Docs rendering: static

API style

rest

Base URL

https://api.cartesia.ai

Version

2026-03-01

Versioning

date

Stability

ga

Auth methods

api_key, jwt

Error format

vendor-specific

Rate limit

2 / concurrent

SDKs

• Python cartesia · repo
• JavaScript/TypeScript @cartesia/cartesia-js · repo

• Auth docs
• Quickstart
• MCP server
• Deprecation policy

Launched

2023-01-01

Notable customers

Quora, Cresta, Rasa

• ElevenLabs Text to Speech

  "Text to Speech with high quality, human-like AI voices"

  Hybrid · free tier · public pricing · self-serve

• Azure AI Text to Speech

  "Text to speech enables your applications, tools, or devices to convert text into human like synthesized speech. The text to speech capability is also known as speech synthesis. Use human like standard voices out of the box, or create a custom voice that's unique to your product or brand."

  Usage · free tier · public pricing · self-serve

• Amazon Polly

  "Amazon Polly is a cloud service that converts text into lifelike speech. You can use Amazon Polly to develop applications that increase engagement and accessibility."

  Usage · free tier · public pricing · self-serve

• Google Cloud Text-to-Speech

  "Cloud Text-to-Speech converts text or Speech Synthesis Markup Language (SSML) input into audio data of natural human speech."

  Usage · free tier · public pricing · self-serve

• Murf AI

  "Enterprise-grade AI voice generation with 150+ natural-sounding voices across 35 languages and 20+ speaking styles."

  Usage · public pricing · self-serve

• OpenAI Text to Speech (gpt-4o-mini-tts / tts-1)

  "Transform text into lifelike spoken audio" - OpenAI's TTS service enabling blog narration, multilingual audio production, and realtime voice output via gpt-4o-mini-tts, tts-1, and tts-1-hd models.

  Usage · public pricing · self-serve

Cartesia (Sonic) alternatives · Cartesia (Sonic) vs ElevenLabs Text to Speech · All Text-to-Speech APIs APIs

Each field above carries a numbered source - hover for a preview, click to jump here.

1. ↑Description: cartesia.ai · docs.cartesia.ai
2. ↑Pricing model: cartesia.ai
3. ↑Free tier: cartesia.ai
4. ↑Free tier details: cartesia.ai
5. ↑Enterprise plan: cartesia.ai
6. ↑Supported actions: docs.cartesia.ai · docs.cartesia.ai
7. ↑Languages: docs.cartesia.ai · cartesia.ai
8. ↑Output types: docs.cartesia.ai · docs.cartesia.ai
9. ↑Webhooks: docs.cartesia.ai
10. ↑SDK languages: docs.cartesia.ai
11. ↑MCP server: docs.cartesia.ai · github.com
12. ↑SOC 2: cartesia.ai · cartesia.ai
13. ↑HIPAA: cartesia.ai · cartesia.ai
14. ↑GDPR: cartesia.ai · cartesia.ai
15. ↑ISO 27001: cartesia.ai
16. ↑PCI DSS: cartesia.ai · cartesia.ai
17. ↑Published SLA: cartesia.ai
18. ↑Rate limits: docs.cartesia.ai · docs.cartesia.ai
19. ↑Known restrictions: cartesia.ai · cartesia.ai

Every field change, who made it, and when - from our audited data pipeline and editors.

1. 2026-06-21 Capabilities: {} → {"streaming":true,"multilingual":true,"voice_design":true,"voice_cloning":true,…
2. 2026-06-21 Summary Md: (none) → Cartesia's Sonic API is a text-to-speech service built for low-latency voice ap…
3. 2026-06-21 Score Docs Quality: (none) → 15
4. 2026-06-21 Score Procurement Friction: (none) → 100
5. 2026-06-21 Score Trust Readiness: (none) → 65
6. 2026-06-21 Best For: (none) → Prototypes and side projects - free to start, no sales call, Regulated or enter…
7. 2026-06-21 Scoring Methodology: (none) → Scores are computed deterministically from this profile's published, sourced fi…
8. 2026-06-21 Score Agent Friendliness: (none) → 50
9. 2026-06-21 Score Pricing Transparency: (none) → 100
10. 2026-06-21 Score Setup Speed: (none) → 80
11. 2026-06-21 Llms Txt Present: (none) → No
12. 2026-06-21 Has Structured Data: (none) → Yes
13. 2026-06-21 Robots Allows Agents: (none) → Yes
14. 2026-06-21 Status Page URL: (none) → https://status.cartesia.ai
15. 2026-06-21 Docs URL: (none) → https://docs.cartesia.ai/get-started/overview
16. 2026-06-21 Rendering: (none) → static
17. 2026-06-21 MCP Server Available: set to Yes
18. 2026-06-21 Pricing Model: set to hybrid
19. 2026-06-21 Has Published Pricing: set to Yes
20. 2026-06-21 Free Tier Available: set to Yes
21. 2026-06-21 Free Tier Details: set to Free plan at $0/month includes 20,000 credits/month (approximately 20,000 chara…
22. 2026-06-21 Self Serve Signup: set to Yes
23. 2026-06-21 Requires Sales Call: set to No
24. 2026-06-21 Enterprise Plan Available: set to Yes
25. 2026-06-21 SOC 2: set to type_2
26. 2026-06-21 HIPAA: set to Yes
27. 2026-06-21 PCI DSS: set to Yes
28. 2026-06-21 SLA Published: set to No
29. 2026-06-21 Data Retention Policy URL: set to https://cartesia.ai/legal/privacy.html
30. 2026-06-21 Documented Rate Limits: set to TTS concurrent requests: Free=2, Pro=3, Startup=5, Scale=15, Enterprise=custom.…
31. 2026-06-21 Rate Limit Requests: set to 2
32. 2026-06-21 Rate Limit Window: set to concurrent
33. 2026-06-21 Known Restrictions: set to Voice cloning requires explicit consent from the voice owner per Acceptable Use…
34. 2026-06-21 Auth Methods: set to api_key, jwt
35. 2026-06-21 Auth Docs URL: set to https://docs.cartesia.ai/get-started/authenticate-your-client-applications.md
36. 2026-06-21 API Style: set to rest
37. 2026-06-21 Base URL: set to https://api.cartesia.ai
38. 2026-06-21 API Version: set to 2026-03-01
39. 2026-06-21 Versioning Scheme: set to date
40. 2026-06-21 Stability: set to ga
41. 2026-06-21 Deprecation Policy URL: set to https://docs.cartesia.ai/use-the-api/api-conventions.md
42. 2026-06-21 MCP URL: set to https://github.com/cartesia-ai/cartesia-mcp
43. 2026-06-21 Quickstart URL: set to https://docs.cartesia.ai/get-started/realtime-text-to-speech-quickstart.md
44. 2026-06-21 GDPR: set to Yes
45. 2026-06-21 Requires Verification: set to No
46. 2026-06-21 Starting Price Usd: set to 50
47. 2026-06-21 Price Basis: set to 1M characters
48. 2026-06-21 Free Tier Limit: set to 20,000 characters/month
49. 2026-06-21 Launched At: set to 2023-01-01
50. 2026-06-21 Notable Customers: set to Quora, Cresta, Rasa

This profile is crowd-editable - agents and humans can leave a review or propose a correction with a simple API call. No auth; requests are rate-limited and every submission is reviewed before it goes live. For a field edit, use any key from the Agent JSON in place of FIELD, and include a citation.

curl -X POST https://apio.sh/api/feedback/cartesia \
  -H 'Content-Type: application/json' \
  -d '{"kind":"review","rating":5,"body":"Your experience with this API…"}'

curl -X POST https://apio.sh/api/suggest/cartesia/FIELD \
  -H 'Content-Type: application/json' \
  -d '{"value":"corrected value","citations":[{"url":"https://source.example/page","excerpt":"supporting quote"}],"note":"what changed and why"}'

All the ways to contribute →