iFLYTEK科大讯飞
Speech recognition and the Spark models
What they do
Founded in Hefei in 1999, iFLYTEK has long specialised in speech recognition and synthesis. It builds the Spark large models, with products in education, healthcare, office hardware and input methods.
Latest coverage
Stories in the feed that mention this company讯飞把语音识别、翻译、情绪理解打包成了一个基座模型，输入法 16.0 就能中英混说，配套的录音环 2 小时录音 2 分钟转完。
星火 X2.6 不靠 N 卡训练，全用国产芯片，还把上下文做到 1M，输入定价 0.9 元/百万 tokens，端侧 4B 模型免费开源，可以试试。
火山引擎靠 45% 份额拿下 IDC 首份 AI 座舱报告的头名，阿里第二，还附带了市场规模和各家打法，做车机方向可以看看。
A 4B model handling a 1M-token context window pushes long-document workloads onto small hardware, making local long-context inference practical on consumer GPUs. The Apache 2.0 license allows unrestricted commercial use, a contrast to models with usage restrictions. Developers building retrieval-free long-context applications can deploy it directly through vLLM.
科大讯飞 Q3 赚了 2.26 亿，但扣非还在亏 1.26 亿，研发一年砸 17 亿赌星火大模型，10 月 22 日还有新版本发布。
长安的启源 Q05 卖得是真不错，连续 6 个月紧凑型 SUV 销冠，新款 13 号就上，7.99 万起的车还内置了讯飞星火大模型。
The model shifts speech recognition from verbatim transcription toward fluent, context-aware text generation, which changes how transcription-heavy workflows like meeting notes and dictation are built. Developers can call it through the iFlytek Open Platform API from 24 Sep, and its stated strength in dialects and noisy scenes is a concrete comparison point against other ASR offerings serving Chinese-language users.
朋友，科大讯飞出了个 AI 录音卡，能自动转文字和总结，还支持 10 种外语翻译，续航 30 小时，比单纯用手机录音好用多了。
For users needing a reliable, multilingual voice model, Spark-Audio-1.0-Preview offers strong performance in complex, noisy environments, comparable to larger, closed-source models.
