[KongchangAI]
Unverified50% confidenceBenchmarkExact time

According to Stridenote, Ollama tests showed that Qwen3.5 35b's speed jumped from 58 token/s to 112 token/s after switching to Apple Silicon's MLX.

1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/10/2026
First Seen
Valid until: 1/8/2027

Sources

Local AI Coding Models Tested: Can 18GB VRAM Beat Opus?

Related Entities

Qwen3.5 35bmodelMLXtechnologyOllamaproductApple Silicontechnology

Related Claims

UnverifiedLing 3.0 Tiny achieves an inference speed of approximately 86–90 tokens per second on a MacBook with M4 Pro, with a peak memory usage of approximately 8.34 GiB under an 8K context.70% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/994076
API
curl https://kongchang.com/api/v1/knowledge/claims/994076
MCP
get_claim(id=994076)