Unverified50% confidenceBenchmarkExact time
On six math word problems, Audio 2B got 3 right in 10 seconds without reasoning enabled, and got 5 right in 68 seconds with reasoning enabled.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/4/2026
First Seen
Valid until: 2/2/2027
Sources
Running Audio and Vision Locally with a 2B Model: LFM3 Hands-On Revealed
bilibili枫叶边城11/4/2026
Related Entities
Related Claims
Unverified对于BERT-base模型推理会话初始化通常耗时数百毫秒，重复执行会使延迟增加1-2个数量级68% similarUnverifiedAfter the model is kept resident, pure text Q&A begins outputting in about 0.38 seconds, and short answers complete in about 0.7 to 1.4 seconds.66% similarUnverifiedIn the author's comparison of three runs each on two tasks, the runs with images totaled 497 seconds and the runs relying only on the control list totaled 381 seconds, all six completed, and relying on the list alone was actually faster.65% similarUnverifiedOfficial data shows that on an entry-level GPU, querying 10 questions together takes a total of 72 milliseconds, averaging approximately 7 milliseconds per question.65% similarUnverifiedThe Anthropic official Python SDK's retry logic defaults to 2 retries, starting at 0.5 seconds and doubling up to 8 seconds, 75%-100% jitter, and honoring the retry-after header; in a 100-concurrency test only 40 succeeded.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1134838API
curl https://kongchang.com/api/v1/knowledge/claims/1134838MCP
get_claim(id=1134838)