IndexTTS Evaluation Results

Model: IndexTTS2 Evaluation Date: 2025/12/08 Paper/Repo: IndexTeam/IndexTTS-2

Metrics Legend:

• WER⬇️: Word Error Rate (lower is better)
• CER⬇️: Character Error Rate (lower is better)
• SIM⬆️: Speaker Similarity (higher is better)

---

Seed-TTS-Eval Benchmark

IndexTTS2

task	dataset	WER/CER⬇️	SIM⬆️	eval_cli	note
tts	seed_tts_eval_en	2.08(1.52)	70.40	[3]
tts	seed_tts_eval_zh	1.04(1.01)	76.04	[4]

---

CV3 Benchmark (Zero-Shot)

IndexTTS2

task	dataset	WER/CER⬇️	SIM⬆️	eval_cli	note
tts	cv3_zero_shot_en	4.32	73.60	[9]
tts	cv3_zero_shot_zh	3.63	78.00	[10]
tts	cv3_zero_shot_hard_en	8.59	74.01	[11]
tts	cv3_zero_shot_hard_zh	8.71	77.21	[12]

---

Evaluation Commands

IndexTTS2

[3] python audio_evals/main.py --dataset seed_tts_eval_en --model indextts2

[4] python audio_evals/main.py --dataset seed_tts_eval_zh --model indextts2

[9] python audio_evals/main.py --dataset cv3_zero_shot_en --model indextts2

[10] python audio_evals/main.py --dataset cv3_zero_shot_zh --model indextts2

[11] python audio_evals/main.py --dataset cv3_zero_shot_hard_en --model indextts2

[12] python audio_evals/main.py --dataset cv3_zero_shot_hard_zh --model indextts2