---
type: "Article"
title: "GLM-5.3-Flash EXL3 4bpw mit TensorFold auf 2× DGX Spark"
description: "GLM-5.3-Flash EXL3 4bpw mit TensorFold auf 2× DGX Spark: Rezept von tonyd2wild mit Tempo, Bedingung und Quelle je Zahl, Gewichten und Lizenz."
resource: "https://www.contextstudios.ai/de/lokale-ki/tonyd2wild--glm-5.3-flash-tensorfold-exl3-4bpw-2x-dgx-spark"
language: "de"
generated:
  by: "process:contextstudios-md/1"
  at: "2026-11-11T03:30:14.088Z"
status: "stable"
---

# GLM-5.3-Flash EXL3 4bpw mit TensorFold auf 2× DGX Spark

GLM-5.3-Flash EXL3 4bpw mit TensorFold auf 2× DGX Spark: 44 tok/s laut github.com (Datensatz-Stand: 3. Nov. 2026).

- Kreator: [tonyd2wild](https://www.contextstudios.ai/de/lokale-ki/kreatoren/tonyd2wild.md)
- Engine: TensorFold v0.6.0 + MiaAI-Lab GLM-5.3-Flash patch series (engine-multi, + tonyd2wild changes)
- Quantisierung: EXL3 4bpw routed experts (mcg) + q4 dense (at load) + FP8 KV
- Modellfamilie: GLM-5.3-Flash
- Hardware: dgx-spark x2
- Kontext: 500000
- Artificial Analysis Intelligence Index: 41.8 ([source](https://artificialanalysis.ai/models/glm-5-3-flash))

## Alle Zahlen

- 44 tok/s (Alltag; Alltag, 1 Anfrage, Prosa-Prompt, mit MTP + DFlash2 (laut Rezept)) — [source](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark/blob/main/runs/2026-10-12-exl3-q4-tp2/TABLES.md#single-stream-decode-toks-median-of-3-time-to-first-token)
- decode_tps: 44 (1 Anfrage) — [source](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark/blob/main/runs/2026-10-12-exl3-q4-tp2/TABLES.md#single-stream-decode-toks-median-of-3-time-to-first-token)
- decode_tps: 77.1 (1 Anfrage) — [source](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark/blob/main/runs/2026-10-12-exl3-q4-tp2/TABLES.md#c1c6-mixed-real-prompts-aggregate-toks-median--peak-of-3-rounds)
- decode_tps: 109.9 (1 Anfrage) — [source](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark/blob/main/runs/2026-10-12-exl3-q4-tp2/TABLES.md#single-stream-decode-toks-median-of-3-time-to-first-token)
- prefill_tps: 1781 (1 Anfrage) — [source](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark/blob/main/runs/2026-10-12-exl3-q4-tp2/TABLES.md#cold-prefill-salted-prompts-one-output-token)
- ttft_ms: 250 (1 Anfrage) — [source](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark/blob/main/runs/2026-10-12-exl3-q4-tp2/TABLES.md#single-stream-decode-toks-median-of-3-time-to-first-token)

## Gewichte

- [brandonmusic/GLM-5.3-Flash-tr3-4bpw](https://huggingface.co/brandonmusic/GLM-5.3-Flash-tr3-4bpw) (main, other)
- [zai-org/GLM-5.3-Flash-BF16](https://huggingface.co/zai-org/GLM-5.3-Flash-BF16) (base, mit)
- [incoai/GLM-5.3-Flash-DFlash2](https://huggingface.co/incoai/GLM-5.3-Flash-DFlash2) (draft, cc-by-nc-nd-4.0)

[https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark](https://github.com/tonyd2wild/GLM-5.3-Flash-TensorFold-DGX-Spark)

## Related

- [Lokale KI auf Ihrer Hardware.](https://www.contextstudios.ai/de/lokale-ki.md)