[KongchangAI]
Unverified50% confidenceSolutionExact time

The system prompt informs the model that it is running on an 80-column, 16-color 286, causing Qwen to proactively keep responses brief and use plain ASCII characters

1
Sources
50%
Confidence
Long-term
Relevance
10/1/2026
First Seen

Sources

Bridging a 40-Year Tech Gap: Running AI Chat and Image Generation on a 286

Related Entities

QwenmodelDeskMindproduct

Related Claims

UnverifiedRunning quantized models from the Qwen3 series on a Mac M3 Ultra, inference speed reaches approximately 34 tokens/s under Q8 quantization.63% similarUnverifiedQwen2.5-7B-Instruct 在多项中文基准测试上的表现接近甚至超过早期 70B 级别的模型61% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/965406
API
curl https://kongchang.com/api/v1/knowledge/claims/965406
MCP
get_claim(id=965406)