As of: 2026-11-13 13:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/m/clover-35b-a3b-preview/

# Clover-35B-A3B-Preview

DiffuSpace, Clover family. Status: preview. Type: reasoning model. Released: 2026-10-28. Training cutoff: not published.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Hugging Face (open weights, BF16 ~69 GB; custom code, trust_remote_code) | `Diffuspace/Clover-35B-A3B-Preview` |  | https://huggingface.co/Diffuspace/Clover-35B-A3B-Preview |
| GitHub (SGLang and llama.cpp inference backends, self-speculative decoding) |  |  | https://github.com/DiffuSpace/Clover-35B-A3B-Preview |
| ModelScope |  |  | https://modelscope.cn/models/Diffuspace/Clover-35B-A3B-Preview |
| Web demo (application required) |  |  | https://chat.diffuspace.ai/ |

## Pricing

- Price not published.
- Last checked: 2026-10-29

## Specs

- Input: text
- Output: text
- Context window: 131,072 tokens
- Max output: not published
- Open weights: yes
- Licence: apache-2.0

## What it has not seen

- No published training cutoff. Counted from its release (2026-10-28): at least 333 AI events since (65 major, 5 historic), as of 2026-11-13.
- Everything since September 2026: https://postcutoff.com/since/2026-09/

## What stands out

- Block-diffusion MoE language model for agents: Text diffusion model derived from Qwen3.6-35B-A3B (35B total, ~3B active, 256 experts top-8): generates 32-token blocks in parallel by unmasking confident positions; DiffuSpace reports SWE-bench Verified 73.6, Terminal-Bench 2.1 51.7 and τ³ 65.3, above Qwen3.6-35B-A3B and the diffusion models LLaDA2.2 flash and Mercury 2.5 (vendor numbers). Source: https://diffuspace.ai/blog/clover-35b/
- Fast single-stream generation: Vendor-measured 680 tokens/s at batch size 1 in BF16 on one GB300 on internal agent workloads, and 515 tokens/s on one H200 vs 365 tokens/s for a DFlash2-accelerated Qwen of the same size. Source: https://diffuspace.ai/blog/clover-35b/

## Notes

First model from DiffuSpace, a company started by members of HKU NLP's Dream diffusion-LM team. Preview release; thinking mode on by default; generation needs the repository's custom code (Transformers 5.19). Long generations can repeat, so the defaults add presence, repetition and length penalties. All benchmarks are the vendor's own; no independent leaderboard results yet.

Open-weight diffusion language models have usually trailed autoregressive ones of the same size on agentic coding; DiffuSpace
claims Clover beats its own autoregressive base (Qwen3.6-35B-A3B) on SWE-bench Verified. Quick start (model card): Transformers
5.19, `trust_remote_code=True`, BF16 on one GPU with ~69 GB plus inference state, `generate(..., block_size=32, threshold=0.95)`.

Model page: https://postcutoff.com/m/clover-35b-a3b-preview/