GPU memory allocation is unverified; external graphics hardware costs extra.
Listing evidence: 2026-09-24. Reatan manufacturer store, US plug and 48GB / 1TB selected, observed 25 Sep 2026 at USD 1299.99. The manufacturer marks this configuration sold out. Dated manufacturer reference only; Amazon price and availability are unverified. Tax and shipping may differ.
Fixed 36GB memory and internal storage; no CUDA support; price is a manufacturer reference rather than a verified Amazon offer.
Listing evidence: 2026-10-02. Apple US launch price for the base M5 Max Mac Studio, published 22 September 2026: USD 2,499. Dated manufacturer reference, not an Amazon offer; tax and shipping may differ.
Memory is soldered and shared with the operating system.
Listing evidence: 2026-09-24. GMKtec, EVO-X2 / 64GB / 1TB, recorded 22 Sep 2026; exact ASIN capacity identified 24 Sep. This is a dated reference, not a current Amazon price; tax, shipping and bundle contents may differ.
Weight size is not tested compatibility or speed. Check runtime support and memory for context before buying. Prices are dated references. As an Amazon Associate I earn from qualifying purchases.
How these options are selected
Up to three platforms, with the lowest dated reference price on each. Unknown prices follow known prices. Each memory pool keeps a reserve: 30% (minimum 2 GB) for system or unified memory; 20% (minimum 1 GB) for dedicated VRAM. System RAM is for CPU inference. This check does not establish GPU allocation, software support or context capacity.
Model file and details
Provider
DiffuSpace
Total parameters
34,661,350,848
Exact file size
22,998,009,248 bytes
Quantization / precision
Q4_K_M
Weight repository
Diffuspace/Clover-35B-A3B-Preview-GGUF
Original model repository
Diffuspace/Clover-35B-A3B-Preview
Weight revision
32e22a02b885e173177b7ce7661ad13448b4f89e
Model revision
4b0402b65e523d0183a14c431955438714cb16ff
Runtime notes
Confirm model architecture and quantization support in your runner.
Recorded release date
2026-10-29
Date evidence
Official GGUF repository creation date; source repository created 24 October. Exact public launch date is not independently established. Official cards and launch blog reviewed 6 November 2026.
These scores were compiled from the linked sources. These published scores were not independently tested here. Different prompts, precision, thinking budgets and tools can change results.
MMLU-Pro
84.8 / 100
Knowledge and reasoning across many subjects. Published accuracy; different evaluation settings can change the result.
Benchmark efficiency: 3.687 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher launch-blog source-model results. Official card recommends its custom SGLang implementation for evaluation; exact benchmark harness revisions, sampling settings and run aggregation are not disclosed. Agent scores depend on external tools.
Publisher-reported source-checkpoint result; not an independent test or a score measured on this Q4_K_M file. No confidence interval is supplied. Download bytes cover language weights, not total runtime memory. Apache-2.0. Text-only block-diffusion MoE; about 3B active parameters but all 34.66B weights are stored. 131,072-token context. Requires Clover-enabled llama.cpp PR #31306 or official custom SGLang, not verified stock-app support. Reported datacenter throughput is not local Q4 speed.
Hard science questions. This is a science reasoning test, not an overall intelligence score.
Benchmark efficiency: 3.644 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher launch-blog source-model results. Official card recommends its custom SGLang implementation for evaluation; exact benchmark harness revisions, sampling settings and run aggregation are not disclosed. Agent scores depend on external tools.
Publisher-reported source-checkpoint result; not an independent test or a score measured on this Q4_K_M file. No confidence interval is supplied. Download bytes cover language weights, not total runtime memory. Apache-2.0. Text-only block-diffusion MoE; about 3B active parameters but all 34.66B weights are stored. 131,072-token context. Requires Clover-enabled llama.cpp PR #31306 or official custom SGLang, not verified stock-app support. Reported datacenter throughput is not local Q4 speed.
Choosing and calling functions. Published BFCL v4 scores; tool environments and settings differ.
Benchmark efficiency: 2.952 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher launch-blog source-model results. Official card recommends its custom SGLang implementation for evaluation; exact benchmark harness revisions, sampling settings and run aggregation are not disclosed. Agent scores depend on external tools.
Publisher-reported source-checkpoint result; not an independent test or a score measured on this Q4_K_M file. No confidence interval is supplied. Download bytes cover language weights, not total runtime memory. Apache-2.0. Text-only block-diffusion MoE; about 3B active parameters but all 34.66B weights are stored. 131,072-token context. Requires Clover-enabled llama.cpp PR #31306 or official custom SGLang, not verified stock-app support. Reported datacenter throughput is not local Q4 speed.
Fixing real software issues with tools. Agent setup and external tools affect these scores. File size covers model weights only.
Benchmark efficiency: 3.2 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher launch-blog source-model results. Official card recommends its custom SGLang implementation for evaluation; exact benchmark harness revisions, sampling settings and run aggregation are not disclosed. Agent scores depend on external tools.
Publisher-reported source-checkpoint result; not an independent test or a score measured on this Q4_K_M file. No confidence interval is supplied. Download bytes cover language weights, not total runtime memory. Apache-2.0. Text-only block-diffusion MoE; about 3B active parameters but all 34.66B weights are stored. 131,072-token context. Requires Clover-enabled llama.cpp PR #31306 or official custom SGLang, not verified stock-app support. Reported datacenter throughput is not local Q4 speed.
Agentic terminal-use task success. Publisher-reported percentages are retained, but agent scaffold, tools, timeouts and reasoning budget can materially affect results.
Benchmark efficiency: 2.248 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher launch-blog source-model results. Official card recommends its custom SGLang implementation for evaluation; exact benchmark harness revisions, sampling settings and run aggregation are not disclosed. Agent scores depend on external tools.
Publisher-reported source-checkpoint result; not an independent test or a score measured on this Q4_K_M file. No confidence interval is supplied. Download bytes cover language weights, not total runtime memory. Apache-2.0. Text-only block-diffusion MoE; about 3B active parameters but all 34.66B weights are stored. 131,072-token context. Requires Clover-enabled llama.cpp PR #31306 or official custom SGLang, not verified stock-app support. Reported datacenter throughput is not local Q4 speed.
Context and version evidence: The repository and file name identify the selected model. Revision hashes are shown above when recorded. Tested context length, full hardware configuration and run date are not consistently recorded in this snapshot. Only settings explicitly stated in each benchmark record are known.
The 22.998 GB figure covers model weights in this file. It excludes the context cache, runtime, operating system and any separate multimodal files. It is not a measured RAM requirement or a token-generation speed.
Confirm model architecture and quantization support in your runner. The selected weights use 22.998 GB; this excludes context cache, runtime buffers and other programs.
These are the first listed tiers that pass our reserve policy, not measured minimum requirements. System RAM and VRAM remain separate. Use the memory tool to change the reserve for your context and runner.