# Clover-35B-A3B-Preview ![Organization](https://img.shields.io/badge/Org-DiffuSpace-9333ea) ![Context](https://img.shields.io/badge/Context-128K%20tokens-1f6feb) ![License](https://img.shields.io/badge/License-Apache%202.0-2da44e) ![Modality](https://img.shields.io/badge/Modality-Text-8957e5) ![Released](https://img.shields.io/badge/Released-Oct%202026-555) > **Quick answer:** Clover-35B-A3B-Preview is DiffuSpace's first release — a 35B-parameter mixture-of-experts **block diffusion** language model (~3B active parameters per token) derived from Qwen3.6-35B-A3B's backbone, with a 128K-token context window and a thinking mode. It scores 73.6% on SWE-bench Verified and 51.7% on Terminal-Bench 2.1, leading the diffusion-model comparison set (LLaDA2.2 flash, Mercury 2.5) and similarly sized autoregressive models (Qwen3.6-35B-A3B, Qwen3.5-35B-A3B) on agentic coding tasks. Open weights under Apache 2.0. ## At a Glance **Where Clover-35B-A3B-Preview leads** - Highest score among compared models on all six agent benchmarks in DiffuSpace's launch comparison: 73.6% SWE-bench Verified, 51.7% Terminal-Bench 2.1 - Strong knowledge/reasoning: 83.8% GPQA Diamond, 95.8% ZebraLogic, 72.0% AA-LCR - Faster generation via parallel block diffusion decoding: DiffuSpace reports 680 tok/s on NVIDIA GB300 and 515 tok/s on H200 for internal agent workloads, versus 365 tok/s for a DFlash2-accelerated same-size Qwen model - Runs INT4-quantized on a 16GB NVIDIA RTX 5070 Ti **Where it lags** - Preview-stage release — DiffuSpace's own comparisons are self-reported, not yet independently reproduced - Text-only (no vision/audio modality) - Block-diffusion decoding is architecturally novel; tooling/ecosystem support (vLLM, inference providers) is less mature than for established autoregressive models **Best for:** agentic coding and tool-use workloads where decoding speed compounds across long agent loops — DiffuSpace positions it directly against similarly sized autoregressive models on exactly this use case. ## What Clover-35B-A3B-Preview Is Clover-35B-A3B-Preview is DiffuSpace's first model, built on the Hugging Face Transformers implementation of Qwen3.6-35B-A3B's hybrid Gated DeltaNet (GDN) / full-attention MoE backbone (40 layers, 256 routed experts, top-8 routing, one shared expert per layer). DiffuSpace redesigned the GDN into a **bidirectional** variant for block diffusion: previously generated context is compressed into the GDN's recurrent state, while tokens within the current block interact through bidirectional softmax attention, letting the model learn when to rely on history versus the current block. Generation proceeds via block diffusion with low-confidence decoding — each block starts fully masked and resolves tokens in parallel as confidence crosses a threshold (default block size 32, threshold 0.95). DiffuSpace's own team traces back to Dream 7B (HKU NLP, April 2025), one of the first diffusion language models shown to compete with autoregressive models; Clover's premise is that the same approach scales to 35B while keeping diffusion's parallel-generation speed advantage. The launch materials emphasize agentic coding and browser/tool-use tasks specifically, citing per-task step counts and wall-clock times (not just benchmark percentages) as evidence that faster generation shortens an agent's full decide-execute-adjust loop. ## Specifications | Field | Value | |---|---| | Organization | DiffuSpace | | Parameters | 35B total, ~3B active per token (MoE) | | Context window | 131,072 tokens (128K) | | Architecture | Bidirectional Gated DeltaNet / full-attention MoE, block diffusion decoding (block_size=32, threshold=0.95); derived from Qwen3.6-35B-A3B | | License | Apache 2.0 | | Release date | October 2026 | | Modality | Text only | ## Pricing Open weights, self-hostable — no per-token API pricing. Available on [Hugging Face](https://huggingface.co/Diffuspace/Clover-35B-A3B-Preview) (~69GB BF16) and [ModelScope](https://modelscope.cn/models/Diffuspace/Clover-35B-A3B-Preview), with inference support for SGLang and llama.cpp. DiffuSpace reports successful INT4 deployment on a single 16GB NVIDIA RTX 5070 Ti. ## Public Benchmark Scores | Benchmark | Score | Source | Date | |---|---|---|---| | [SWE-bench Verified](/benchmarks/swe-bench-verified) | 73.6 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [Terminal-Bench 2.1](/benchmarks/tbench/terminalbench-21) | 51.7 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [GPQA Diamond](/benchmarks/rein-et-al/gpqa-diamond) | 83.8 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [ZebraLogic](/benchmarks/bill-yuchen-lin-et-al-allen-institute-for-ai-university-of-washington/zebralogic) | 95.8 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [AA-LCR](/benchmarks/artificial-analysis/aa-lcr) | 72.0 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [BFCL-v4](/benchmarks/patil-et-al-uc-berkeley/bfcl-v4) | 67.9 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [τ³ Banking](/benchmarks/sierra-research/3-banking) | 14.4 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | | [MMLU-Pro](/benchmarks/mmlu-pro) | 84.8 | [DiffuSpace launch blog](https://diffuspace.ai/blog/clover-35b/) | 2026-10 | *Scores above are reported by DiffuSpace and shown for context. They are not Benchgen measurements. DiffuSpace's own comparison set includes LLaDA2.2 flash, Mercury 2.5, Qwen3.6-35B-A3B, Qwen3.5-35B-A3B, MiniCPM-V 4.7, LFM2.5-24B-A2B, Nemotron 3.5 Lightning 30B-A3B, and Gemma 4 26B-A4B across different subsets of this table — see the source blog for full per-model methodology.* **Excluded from the table above (bucket-C, documented per skill procedure):** - **τ³ Overall** (65.3) — a generic/unqualified "τ³" score distinct from our domain-specific `tau3-banking` entity; per standing policy we never silently narrow or widen scope to map a generic name onto a domain-specific entity (same exclusion pattern as prior cycles). The blog separately reports **τ³ Banking** at 14.4, which maps cleanly and is included above. - **MATH-500 generation speed** (101.8 tok/s on RTX 5070 Ti INT4) — a throughput/speed measurement, not an accuracy benchmark score; not a leaderboard-table candidate. - Internal agent-workload throughput figures (680 tok/s GB300, 515 tok/s H200) and UAV-dispatch / WebMCP / route-planning demo timings — proprietary internal task evaluations with no independently reproducible public benchmark definition. - **Decision model** (joint vs. independent answering, 95.0%/80.4% on a 500-group/2,420-question internal validation set) — a separate, not-yet-released DiffuSpace model variant evaluated on an internal dataset; no public spec. ## Related Models & Benchmarks - Compare: [Qwen3.6-35B-A3B](/models/alibaba/qwen3-6-35b-a3b) - Benchmarks: [SWE-bench Verified](/benchmarks/swe-bench-verified) · [Terminal-Bench 2.1](/benchmarks/tbench/terminalbench-21) · [GPQA Diamond](/benchmarks/rein-et-al/gpqa-diamond) ## Frequently Asked Questions

What is Clover-35B-A3B-Preview?

Clover-35B-A3B-Preview is DiffuSpace's first model — a 35B-parameter MoE (~3B active) block-diffusion language model built on a Qwen3.6-35B-A3B-derived backbone, designed for fast agentic coding and tool use.

What is Clover-35B-A3B-Preview's context window?

131,072 tokens (128K) for combined prompt and generated response.

Is Clover-35B-A3B-Preview open source?

Yes — released under the Apache 2.0 license, with weights on Hugging Face and ModelScope.

What makes Clover-35B-A3B-Preview different from a typical LLM?

It uses block diffusion instead of autoregressive token-by-token decoding: each block of tokens starts masked and is resolved in parallel as confidence rises, which DiffuSpace reports gives it faster generation than similarly sized autoregressive models in agentic workloads.

--- *Specs and scores sourced from DiffuSpace's official launch blog (diffuspace.ai/blog/clover-35b) and Hugging Face model card. Last updated 2026-10-31.*

Clover-35B-A3B-Preview — Evaluations

Evaluations