Please don't use that garbage. Just use the base Qwen models or Nex/Orinth, as those are the only properly post-trained finetunes. The Qwopus models are marketing.
You're late to the party, mate; we've been doing this for years. Grab a SearXNG instance, stand up an MCP server for it, and expose the tool into your system prompt. Or use Brave Search. Or Exa if you want to pay. Any of them work. The model will pick it up straight away.
Even llama.cpp's bundled web UI handles it fine. Dead simple.
Did you even use it? It was nerfed to hell and back. It stopped following instructions, forgot what sub-agents responded and so on. Stop spreading this pro-Anthropic narrative. They did a rug pull due to lack of compute.
A3B-35B is better suited for laptops with enough VRAM/RAM.
This dense model however will be bandwidth limited on most cards.
The 5090RTX mobile sits at 896GB/s, as opposed to the 1.8TB/s of the 5090 desktop and most mobile chips have way smaller bandwith than that, so speeds won't be incredible across the board like with Desktop computers.
For autocomplete, Qwen 3.5 9B should be enough even at Q4_k_m.
The upcoming coding/math Omnicoder-2 finetune might be useful (should be released in a few days).
Either that or just load up Qwen3.5-35B-A3B-Q4_K_S
I'm serving it at about 40-50t/s on a 4070RTX Super 12GB + 64GB of RAM. The weights are 20.7GB + KV Cache (which should be lowered soon with the upcoming addition of TurboQuant).