Unverified50% confidenceReleaseExact time
Google has released a small Gemma-series model targeting speech transcription scenarios (labeled as Diarization Gemma 4 E4B related), designed for speaker diarization and error correction, with a total weight of approximately 8B parameters.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/8/2026
First Seen
Valid until: 1/6/2027
Sources
Related Entities
Related Claims
UnverifiedWisprGemma 使用单个 Gemma E2B 模型在单次推理中完成语音识别、文本清理和重写69% similarUnverifiedGemma 3 is an open-source multimodal language model series released by Google DeepMind that can process both text and image inputs.69% similarUnverifiedEmbeddingGemma 2 is based on the Gemma 3 decoder architecture, shares the text tokenizer and audio encoder with Gemma 3, has a context window of 8K tokens (4 times that of the previous generation), and outputs vectors with a dimension of 768.66% similarUnverifiedEmbeddingGemma 2 is built on Gemma 4 and shares its text tokenizer and audio encoder, enabling both models to run simultaneously within a unified pipeline with lower total memory usage.63% similarUnverifiedGemma 4 eliminates the traditional encoder structure, enabling small-sized models to run deep thinking tasks on consumer-grade GPUs.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/987117API
curl https://kongchang.com/api/v1/knowledge/claims/987117MCP
get_claim(id=987117)