<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>GPUfit — GeForce GTX 1660 Ti</title><link>https://gpufit.tech/gpu/gtx-1660-ti</link><description>New catalog models scored against GeForce GTX 1660 Ti.</description><item><title>bottlecapai ThinkingCap gemma 4 26B A4B it on GeForce GTX 1660 Ti: Expert offload</title><link>https://gpufit.tech/models/bottlecapai-thinkingcap-gemma-4-26b-a4b-it</link><guid>https://gpufit.tech/models/bottlecapai-thinkingcap-gemma-4-26b-a4b-it</guid><pubDate>Fri, 23 Oct 2026 21:26:38 GMT</pubDate><description>Expert offload on your GeForce GTX 1660 Ti: Q3_K_M keeps shared weights on the GPU and about 8.77 GB of experts in system RAM. About 58–79 tok/s. Active-parameter speed is approximate. 58–79 tok/s.</description></item><item><title>NVIDIA Nemotron 3.5 Bolt E4B on GeForce GTX 1660 Ti: Runs</title><link>https://gpufit.tech/models/nvidia-nemotron-3-5-bolt-e4b</link><guid>https://gpufit.tech/models/nvidia-nemotron-3-5-bolt-e4b</guid><pubDate>Fri, 23 Oct 2026 09:47:55 GMT</pubDate><description>Runs fully on your GeForce GTX 1660 Ti at UD-Q3_K_XL, 8K context, about 32–44 tok/s. Weights 3.93 GB plus 0.14 GB of KV cache — 4.41 GB of 5.22 GB usable. Up to 32K context still fits. 32–44 tok/s.</description></item><item><title>DeepSeek V4.1 Flash 1017 on GeForce GTX 1660 Ti: Won't fit</title><link>https://gpufit.tech/models/deepseek-v4-1-flash-1017</link><guid>https://gpufit.tech/models/deepseek-v4-1-flash-1017</guid><pubDate>Sat, 17 Oct 2026 20:36:06 GMT</pubDate><description>Won't fit on GeForce GTX 1660 Ti. The smallest quant UD-IQ1_S is 217.0 GB before KV cache, and usable memory is 5.22 GB.</description></item><item><title>LFM2.5 24B A2B on GeForce GTX 1660 Ti: Expert offload</title><link>https://gpufit.tech/models/lfm2-5-24b-a2b</link><guid>https://gpufit.tech/models/lfm2-5-24b-a2b</guid><pubDate>Tue, 13 Oct 2026 14:19:48 GMT</pubDate><description>Expert offload on your GeForce GTX 1660 Ti: UD-Q4_K_XL keeps shared weights on the GPU and about 11.5 GB of experts in system RAM. About 83–113 tok/s. Active-parameter speed is approximate. 83–113 tok/s.</description></item></channel></rss>