AI Agent StoreDiffuSpace · Download verified
Clover 35B A3B Preview
Try a downloadable coding and reasoning model on your own hardware. This preview needs its supplied serving code and substantial memory.
Released 2026-10-30 · First reviewed 2026-10-31 · Checked 2026-11-03.
We still need comparable independent tests. You can compare the confirmed features and limits below, but we cannot yet give this version an Agent Fit or Practical Value rating. Try it on your hardware and tasks before relying on it.
What you can use today
You download the model files and supply the hardware and software to run them. Capacity is the amount of information a request can contain; settings and available memory can limit what fits.
Exact download ID: Diffuspace/Clover-35B-A3B-Preview. A text-only diffusion model derived from Qwen3.6-35B-A3B, with 35 billion total parameters and about 3 billion active per step. It supports thinking and tool-call formatting. Publisher speed and coding results are not independent measurements here.
Apache 2.0 weights with upstream attribution requirements. The supplied Transformers implementation uses custom repository code, Transformers 5.19.0 and PyTorch 2.10 or newer. BF16 weights alone are about 69 GB; running the model needs additional memory. The publisher’s lighter demos do not establish a universal hardware requirement. Separate SGLang and llama.cpp setup guides are still promised.
The 131,072-token limit includes the prompt and generated answer. Generation reserves complete blocks, so leave room for the final block. The example output budget is not a verified maximum.
No verified hosted token price. Compare hardware and serving costs using your actual workload.
Provider release and specifications ↗
Check availability and plans ↗
More setup and licensing details ↗
More setup and licensing details ↗
Test it before switching
Give each model the same examples, tools and definition of success. Record correct outcomes, response time, retries and total cost. Include the time and hardware needed to run it.
Compare caching and long-context costs ↗