Skip to content
IFAI

Model rankings

906 models · 5 boards · 2 sources

Text-to-speech arena · Anonymous head-to-head votes · Updated January 3, 2026

Best voice
Inworld TTS 1 MaxInworld1,172Score
Best open model
Kokoro 82M v1.0Kokoro1,059Score
Best from China
Speech 2.6 HDMiniMax1,153Score
Top three by Score
1Inworld TTS 1 MaxInworld1,172
2Speech 2.6 HDMiniMax1,153
3Speech 2.6 TurboMiniMax1,147
Top three open models
1Kokoro 82M v1.0Kokoro · #15 overall1,059
2Fish Speech 1.5Fish Audio · #30 overall1,029
3Maya1Maya Research · #31 overall1,027
This week's changes
Speech 2.6 HDSpeech · #2New
Speech 2.6 TurboSpeech · #3New
minimax-m2.1-previewCoding · #5New
Multilingual v2Speech · #7New
GLM-4.7 (Reasoning)Intelligence · #8New
RankChangeModelScore95% interval
1—TTS-1ClosedOpenAI1,107—
2▲ 2Polly GenerativeClosedAmazon1,062—
3▼ 1TTS-1 HDClosedOpenAI1,058—
4▲ 1Azure NeuralClosedMicrosoft Azure1,051—
5▼ 2Polly Long-FormClosedAmazon1,049—
6—GPT-4o mini TTSClosedOpenAI1,046—
6▲ 1StudioClosedGoogle1,046—
8▲ 1JourneyClosedGoogle1,038—
9▲ 1Chirp 3: HDClosedGoogle1,025—
10▼ 2Magpie MultilingualClosedNVIDIA1,002—
11—VibeVoice 1.5BClosedMicrosoft Azure992—
12—Polly NeuralClosedAmazon935—
13—WaveNetClosedGoogle905—
14—StandardClosedGoogle875—
15—Neural2ClosedGoogle873—
16—Polly StandardClosedAmazon847—

Intelligence over time

Each maker's best model at the time on the Artificial Analysis Intelligence Index, by the month it was released.

Gained most · Anthropic +57.2Highest now · Anthropic 64.6Gap between highest and lowest 18.1 ptsTime range Nov 2025 – Nov 2026
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi

Arena data comes from the leaderboard dataset it publishes on Hugging Face (CC BY 4.0). Only each board's overall leaderboard is used; scores and ranks are unchanged. dataset · CC BY 4.0

Artificial Analysis data comes from its API (artificialanalysis.ai). Scores are unchanged; ranks follow the scores. artificialanalysis.ai

Open or closed follows the license Arena gives; prices and Artificial Analysis intelligence show only where model names match, otherwise “—”.