Best local LLMs for the Intel Xeon D-1600 series NAS (8 GB)

Each list assumes the context that task needs, so a model that fits for chat may rank lower for long documents.

General chat at 8K context

  1. 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 E2B Q4_K_M · 12.8 tok/s · Runs well
  2. 50Fair quality (C) · 50 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. Ling 3.0 Tiny (8B-A1.3B) Q3_K_M · 26.8 tok/s · Runs well (tight fit)
  3. 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 2B Q4_K_M · 11.7 tok/s · Runs well
  4. 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Bolt E4B Q2_K · 10.8 tok/s · Runs well
  5. 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Nanbeige4.2 3B Q2_K · 11.6 tok/s · Runs well

Coding at 32K context

  1. 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 2B Q4_K_M · 11.7 tok/s · Runs well
  2. 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 E2B Q4_K_M · 12.8 tok/s · Runs well
  3. 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q2_K · 11.8 tok/s · Runs well
  4. 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. Nemotron Labs Dory 1.0 4B Q2_K · 12.2 tok/s · Runs well (tight fit)
  5. 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. Unlimited OCR 2 (3B-A0.6B) Q4_K_M · 42.1 tok/s · Runs well (tight fit)

Reasoning at 32K context

  1. 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 2B Q4_K_M · 11.7 tok/s · Runs well
  2. 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 E2B Q4_K_M · 12.8 tok/s · Runs well
  3. 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q2_K · 11.8 tok/s · Runs well
  4. 47Fair quality (C) · 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. Nemotron Labs Dory 1.0 4B Q2_K · 12.2 tok/s · Runs well (tight fit)
  5. 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. Unlimited OCR 2 (3B-A0.6B) Q4_K_M · 42.1 tok/s · Runs well (tight fit)

Documents (RAG) at 32K context

  1. 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 E2B Q4_K_M · 12.8 tok/s · Runs well
  2. 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. Unlimited OCR 2 (3B-A0.6B) Q4_K_M · 42.1 tok/s · Runs well (tight fit)
  3. 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. Unlimited OCR (3B-A0.6B) Q4_K_M · 42.1 tok/s · Runs well (tight fit)
  4. 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q4_K_M · 17.0 tok/s · Runs well
  5. 36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Granite 4.0 H Tiny (7B-A1B) Q3_K_M · 33.9 tok/s · Runs well (tight fit)

Agents & tools at 64K context

  1. 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q4_K_M · 17.0 tok/s · Runs well
  2. 50Fair quality (C) · 50 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 2B Q3_K_M · 15.1 tok/s · Runs well (tight fit)
  3. 36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Granite 4.0 H Tiny (7B-A1B) Q3_K_M · 33.9 tok/s · Runs well (tight fit)
  4. 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 E2B Q4_K_M · 12.8 tok/s · Runs well (tight fit)
  5. 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q2_K · 11.8 tok/s · Runs well (tight fit)

Reverse engineering at 32K context

  1. 58Good quality (B) · 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 E2B Q4_K_M · 12.8 tok/s · Runs well
  2. 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 2B Q4_K_M · 11.7 tok/s · Runs well
  3. 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. Unlimited OCR 2 (3B-A0.6B) Q4_K_M · 42.1 tok/s · Runs well (tight fit)
  4. 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. Ling 3.0 Tiny (8B-A1.3B) Q2_K · 33.3 tok/s · Runs well (tight fit)
  5. 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. Unlimited OCR (3B-A0.6B) Q4_K_M · 42.1 tok/s · Runs well (tight fit)

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.

Was this page helpful?