Best local LLMs for the Arduino UNO Q (4 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. 55Good quality (B) · 55 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q6_K · 1.2 tok/s · Very slow
  2. 57Good quality (B) · 57 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. MiniCPM5 2B Q4_K_M · 1.2 tok/s · Very slow (tight fit)
  3. 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. LFM3-VL 2B Q5_K_M · 1.3 tok/s · Very slow
  4. 46Fair quality (C) · 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q2_K · 1.2 tok/s · Very slow (tight fit)
  5. 42Fair quality (C) · 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. TMax 4B Q2_K · 1.1 tok/s · Very slow (tight fit)

Coding at 32K context

  1. 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q4_K_M · 1.7 tok/s · Very slow
  2. 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. LFM3-VL 2B Q5_K_M · 1.3 tok/s · Very slow (tight fit)
  3. 30Basic quality (D) · 30 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 1B Q3_K_M · 3.5 tok/s · Slow
  4. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. PaddleOCR-VL 1.6 0.9B Q4_K_M · 3.1 tok/s · Slow
  5. 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. TMax 2B Q4_K_M · 1.6 tok/s · Very slow

Reasoning at 32K context

  1. 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q4_K_M · 1.7 tok/s · Very slow
  2. 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. LFM3-VL 2B Q5_K_M · 1.3 tok/s · Very slow (tight fit)
  3. 36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. Jamba Reasoning 3B Q3_K_M · 1.2 tok/s · Very slow (tight fit)
  4. 30Basic quality (D) · 30 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. MiniCPM5 1B Q3_K_M · 3.5 tok/s · Slow
  5. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. PaddleOCR-VL 1.6 0.9B Q4_K_M · 3.1 tok/s · Slow

Documents (RAG) at 32K context

  1. 53Fair quality (C) · 53 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. LFM3 1.7B Q4_K_M · 1.7 tok/s · Very slow
  2. 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. LFM3-VL 2B Q5_K_M · 1.3 tok/s · Very slow (tight fit)
  3. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. PaddleOCR-VL 1.6 0.9B Q4_K_M · 3.1 tok/s · Slow
  4. 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. TMax 2B Q4_K_M · 1.6 tok/s · Very slow
  5. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. Jamba2 3B Q4_K_M · 1.0 tok/s · Very slow (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 71/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q4_K_M · 1.7 tok/s · Very slow (tight fit)
  2. 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. TMax 2B Q4_K_M · 1.6 tok/s · Very slow (tight fit)
  3. 42Fair quality (C) · 42 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. LFM3-VL 2B Q3_K_M · 2.0 tok/s · Very slow (tight fit)
  4. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. PaddleOCR-VL 1.6 0.9B Q4_K_M · 3.1 tok/s · Slow (tight fit)
  5. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. Jamba2 3B Q4_K_M · 1.0 tok/s · Very slow (tight fit)

Reverse engineering at 32K context

  1. 53Fair quality (C) · 53 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. LFM3 1.7B Q4_K_M · 1.7 tok/s · Very slow
  2. 48Fair quality (C) · 48 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. LFM3-VL 2B Q5_K_M · 1.3 tok/s · Very slow (tight fit)
  3. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. PaddleOCR-VL 1.6 0.9B Q4_K_M · 3.1 tok/s · Slow
  4. 38Basic quality (D) · 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. TMax 2B Q4_K_M · 1.6 tok/s · Very slow
  5. 39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. Jamba2 3B Q4_K_M · 1.0 tok/s · Very slow (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?