WisprFlow 自研语音模型 Canto 1：真实听写错误率 3.4%
语音输入工具 WisprFlow 从租别人的引擎改成自己训模型，还挖来 Alexa 的创始科学家带队。Canto 1 在 2300 多个真实用户的嘈杂听写上跑出 3.4% 词错率，比测试过的竞品都低。做语音产品或关心 ASR 的开发者可以看看它的 RL 数据闭环思路。
WisprFlow 聘请前 Alexa 创始科学家 Ariya Rastrow 担任首席科学官，放弃租用第三方语音引擎，转向自研模型。其首个模型 Canto 1 的目标是减少用户对听写文本的编辑次数，训练数据来自嘈杂的真实听写场景，而非传统的清晰朗读音频。
在超过 2,300 名说话者的约 10 小时真实 Flow 听写数据上，Canto 1 取得 3.4% 的词错率，低于 Wispr 测试的所有竞品模型。自发布以来，出现时间超过 1 秒的听写减少了 54%——对语音输入来说，反馈速度和准确率同样关键，因为用户要等整句结束才能确认机器听到了什么。
训练流程沿用大语言模型的三阶段：预训练、监督微调和基于 GRPO 的强化学习。RL 阶段使用自愿共享数据的 Flow 用户纠错：Wispr 会比对音频与文本的对齐情况，把误听词纳入训练参考，而普通句子改写则被忽略。
The keyboard was a workaround for machines that couldn’t hear. Machines that hear make it optional.
WisprFlow recently hired Ariya Rastrow (founding Alexa scientist) as chief science officer and moved from rented speech engines to models it trains itself.
Hard to find a better fit for a speech lab than someone who scaled Alexa’s speech team.
Canto 1, his lab’s first model, was built to cut how often people edit dictated text, so Wispr trained it on noisy real-world dictations and tuned it to each user’s vocabulary and names.
And this is a hard test, unlike many, because Speech recognition was long tested on clean, read-aloud audio, but the hard problem is a cheap laptop mic in a room where someone else is talking.
On about 10 hours of real Flow dictations from more than 2,300 speakers, it posted a 3.4% word error rate, lower than all competing models that Wispr tested.
Typing gives you feedback on every keystroke, while speech makes you wait until the end of a sentence to see what the machine heard, so speed counts as much as accuracy.
Since launch, dictations that take more than 1 second to appear are down 54%.
Its training runs through pretraining, supervised fine-tuning and reinforcement learning with GRPO, the same stages used for large language models.
The reinforcement learning stage uses corrections from Flow users who opted in to share data.
When someone fixes a dictated word, Wispr checks how the audio aligns with the text and how the edit is shaped.
A misheard word gets folded back into the training reference, and a plain rewrite of the sentence gets ignored.
本条情报归属 AITOP「模型」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」「中美大模型技术代差与商业化对照」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：WisprFlow、Canto 1、语音识别、GRPO，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「rohanpaul_ai」（@rohanpaul_ai），经 AITOP 多维评分管线结构化提炼。
