AiRecMark/AI Tools/Audio & Music/Kokoro TTS
INDEX / AUDIO_SPEECH / KOKORO_TTS
huggingface.co verified
KO

Kokoro TTS

Verified Benchmark

Open-weight 82M-parameter TTS model with premium voice quality

FREE PRICING SCORE 82.5/100 4 RECORDED FEATURES UPDATED 2026-09-17
AiRecMark Score
82.5 /100
trending_up v2-5dim
workspace_premium #11 in Voice & Music AI
Launch Console north_east
Output Quality record_voice_over
84%

Quality & reliability of primary audio output.

Feature Depth bolt
70%

82M open-weight model

Usability fingerprint
74%

Onboarding & editor ergonomics.

Entry Price payments
Open-weight (Apache 2.0)

Open-weight (Apache 2.0) · hosted API serving under $1 per M chars

Scoring Vectors

Deterministic evaluation across the AiRecMark five-dimension rubric (v2-5dim)

SNAPSHOT 2026-09-17
sentiment_satisfied Output Quality & Reliability (Weighted 25%) 84%
format_quote Feature Depth & Integrations (Weighted 20%) 70%
noise_aware Onboarding & Usability (Weighted 15%) 74%
auto_stories Runtime Performance & Latency (Weighted 20%) 86%
graphic_eq Price-to-Value Efficiency (Weighted 20%) 96%
PACKET TRANSIT DISTRIBUTION (PING JITTER) 5-dim spread ±13.0 pts (5-dim spread)
40ms 65ms (Median) 75ms (Flash TTFT) 120ms
graphic_eq ARCHIVE RECORD
v2-5dim
QUALITY DIM 84%
FEATURES DIM 70%
VALUE DIM 96%
terminal quickstart.py 
# Kokoro TTS — official documentation

"https://huggingface.co"


# Pricing source (T0): https://huggingface.co/hexgrad/Kokoro-82M

# Entry tier: Open-weight (Apache 2.0) · hosted API serving under $1 per M chars (checked 2026-09-17)

# Rubric: v2-5dim · recorded features: 4

Quantitative Technical Specification

Core architectural subsystems in production release v2.5

speed

82M open-weight model

Kokoro-82M v1.0 released January 2025 under Apache 2.0.

Capability T0 Verified
mic_double

8 languages, 54 voices

Voicepacks shipping with the v1.0 release.

Capability T0 Verified
support_agent

StyleTTS 2 decoding

StyleTTS 2-style generator without diffusion for speed.

Capability T0 Verified
surround_sound

Cheap serving

Documented API market rate under $1 per million characters.

Capability T0 Verified

Direct Peer Matrix: Voice Synthesis & Conversational Engines

Standardized comparative evaluation using normalized 1,000-character test payloads

Benchmarks Updated 2026-09-14
Model & Provider AiRecMark Score Pricing Base Performance (rubric) Languages Prosody Fidelity
Kokoro TTS Leader 82.5 Open-weight (Apache 2.0) 86/100 — 84%
ElevenLabs 95.4 Usage API / Tiered — — —
PlayHT 89.2 Freemium — — —
Cartesia 85.8 $5/mo — — —
Adobe Podcast 85.6 $0/mo — — —

Commercial Tiers & Compute Allocation

Transparent character pools, concurrent socket capacities, and API rate limits

Pro Popular
Custom /mo

Open-weight (Apache 2.0) · hosted API serving under $1 per M chars

  • check 82M open-weight model
  • check 8 languages, 54 voices
  • check StyleTTS 2 decoding
Enterprise
Custom SLA terms

Dedicated infrastructure, security review, and compliance support.

  • check SSO / SAML & audit controls
  • check Dedicated support channel
  • check Custom quota & SLA
verified_user
AIrecmark Analyst Consensus Verdict SCORING MODEL: v2-5dim

Kokoro TTS stands out for apache 2.0 open weights. Open-weight 82M-parameter TTS model with premium voice quality anchors its proposition, and AiRecMark's five-dimension audit lands it at 82.5/100 — a pragmatic default for developers Embedding Low-Cost TTS.

Lead Evaluator: AiRecMark Audio Analyst Group
ARCHIVE SNAPSHOT: 2026-09-17 · huggingface.co