BaiLian

Cosyvoice V3 Plus

AlibabaToken-based
Alias:cosyvoice-v3-plus
Create API Key

Compare cosyvoice-v3-plus API pricing, supported endpoints, capabilities and access options on Modelsell.

tts语音合成streaming声音复刻
Starting price
Input / Output · 1M
Context
—
Maximum input window
Modalities
→
Released
Sep 2025

Pricing by Supplier

Alibaba
阿里巴巴百炼官方
Input$29.4118/ 1M
Output$0/ 1M

Capabilities / Supported modalities

Streaming
Input
Output

Provider & data privacy

Provider
Unknown
Tokenizer
BPE (vendor-specific)
License
Provider-specificUnknown
Data retention54 daysNot used for upstream training by default

Performance

Benchmarks

Scores on standardized evaluations. Higher percentages are better — and rank percentile shows

Metrics sourced fromArtificial Analysis 2026-09-30·Cosyvoice V3 Plus

This model is not included in the current benchmark snapshot.

Missing models or measurements are not zero scores.

Benchmark charts preserve the source model selection and reasoning settings. Missing models or measurements are not zero scores, and benchmark cost or speed is not this site’s service commitment.

About Cosyvoice V3 Plus

把熟悉的声音用于新文稿

CosyVoice-v3-Plus 适合已有明确声音形象、又需要持续制作新内容的项目,例如课程讲师配音、品牌解说和有声读物。可以先用参考音频创建复刻音色,再以相应 voice ID 合成不同正文。模型侧重音质、声音相似度与表现力,支持文本到语音的实时流式生成。

创建复刻音色时,可准备 5—20 秒参考录音。选择说话人单一、吐字完整且没有明显背景音乐的片段,有助于表达清楚的声音特征。先用包含日常短句、数字和专有名词的内容试听,确认声音方向和读音后,再投入长篇制作。音色创建与正文合成是两个步骤,应保存与目标模型匹配的 voice ID。

按音色类型使用控制能力

v3-Plus 的声音复刻和声音设计音色不支持指令控制;系统音色需要按该音色支持的固定格式设置指令。不要把任意表演要求塞进自定义音色请求,期待它自动改变情绪。对自定义声音,可以先通过正文标点、分句和内容编排改善朗读节奏,再试听调整。

长篇有声内容建议按自然段或章节处理,统一人名、缩写和数字写法,检查段落之间的停顿与听感。流式场景可逐段接收音频;完整文件场景则应及时保存合成结果,避免仅保留有时效的下载链接。

北京地域基准价为 2 元/万计费字符,输出音频不另收费。汉字通常计 2 个字符,SSML 标签不计入,按返回字符数结算。美元基准按 1 美元 = 6.8 元换算,最终费用应用账户分组倍率。

Use cases and prompting

制作讲师声音的课程片段

用清晰的 5—20 秒参考录音完成音色创建,取得适配 cosyvoice-v3-plus 的 voice ID。

待朗读正文:“今天我们先认识光的反射。请观察镜子中的光线,再思考入射角与反射角之间的关系。”

为什么自定义音色不响应情绪指令?

v3-Plus 的复刻与设计音色不支持指令控制。先调整正文分句和标点;系统音色则只使用其支持的固定指令。

录音很短,能直接生成整本书吗?

创建音色后可分段合成正文,但应先做短篇试听,确认专有名词、读音和长句节奏,再制作完整内容。

API access

API documentation

Authentication

All requests must include Authorization: Bearer <TOKEN> header. Anthropic-formatted endpoints accept the x-api-key header instead.

Generate tokens from the Tokens page; you can scope them to specific models, groups, IPs, and rate-limits.

Rate limits

SupplierRPMTPMRPD
AlibabaUnlimitedUnlimitedUnlimited

No restriction

Frequently asked questions about cosyvoice-v3-plus

What is cosyvoice-v3-plus?

Compare cosyvoice-v3-plus API pricing, supported endpoints, capabilities and access options on Modelsell.

How do I call cosyvoice-v3-plus?

Create an API key with access to cosyvoice-v3-plus, then use the exact model ID and a supported endpoint from the API access section. Request fields depend on the selected endpoint.

How is cosyvoice-v3-plus priced?

Pricing depends on the selected provider group and the model billing unit. The current input, output, request, or media prices are shown on this page before sign-up.

How should I evaluate cosyvoice-v3-plus for my project?

Start with the use cases and prompting guidance on this page, then evaluate the model with representative inputs from your project.