Best local LLMs for the Orange Pi AIpro (20T) (12 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. 78Great quality (A) · 78 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. Gemma 4 12B Q4_K_M · 1.5 tok/s · Very slow (tight fit)
  2. 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. SenseNova Loop 9B Preview Q4_K_M · 1.8 tok/s · Very slow
  3. 73Great quality (A) · 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. Ornith 1.5 9B Q4_K_M · 1.8 tok/s · Very slow
  4. 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.5 9B Q4_K_M · 1.8 tok/s · Very slow
  5. 71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. Ornith 1.0 9B Q4_K_M · 1.8 tok/s · Very slow

Coding at 32K context

  1. 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. SenseNova Loop 9B Preview Q4_K_M · 1.8 tok/s · Very slow (tight fit)
  2. 67Good quality (B) · 67 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Nemotron Labs Dory 1.0 4B Q6_K · 3.0 tok/s · Slow
  3. 64Good quality (B) · 64 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. Nanbeige4.2 3B Q4_K_M · 4.2 tok/s · Slow (tight fit)
  4. 64Good quality (B) · 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q5_K_M · 3.5 tok/s · Slow
  5. 73Great quality (A) · 73 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. Ornith 1.5 9B Q4_K_M · 1.8 tok/s · Very slow

Reasoning at 32K context

  1. 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.5 9B Q4_K_M · 1.8 tok/s · Very slow
  2. 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. SenseNova Loop 9B Preview Q4_K_M · 1.8 tok/s · Very slow (tight fit)
  3. 67Good quality (B) · 67 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Nemotron Labs Dory 1.0 4B Q6_K · 3.0 tok/s · Slow
  4. 64Good quality (B) · 64 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. Nanbeige4.2 3B Q4_K_M · 4.2 tok/s · Slow (tight fit)
  5. 64Good quality (B) · 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q5_K_M · 3.5 tok/s · Slow

Documents (RAG) at 32K context

  1. 73Great quality (A) · 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. Ornith 1.5 9B Q4_K_M · 1.8 tok/s · Very slow
  2. 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.5 9B Q4_K_M · 1.8 tok/s · Very slow
  3. 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. SenseNova Loop 9B Preview Q4_K_M · 1.8 tok/s · Very slow (tight fit)
  4. 71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. Ornith 1.0 9B Q4_K_M · 1.8 tok/s · Very slow
  5. 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Bolt E4B Q5_K_M · 3.2 tok/s · Slow

Agents & tools at 64K context

  1. 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Bolt E4B Q5_K_M · 3.2 tok/s · Slow
  2. 64Good quality (B) · 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. Spark-X2.6 4B Q5_K_M · 3.5 tok/s · Slow
  3. 59Good quality (B) · 59 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. TMax 4B Q5_K_M · 3.4 tok/s · Slow
  4. 58Good quality (B) · 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. Agents-A1 4B Q5_K_M · 3.2 tok/s · Slow
  5. 49Fair quality (C) · 49 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. LFM2.5 8B-A1B Q4_K_M · 11.2 tok/s · Runs well

Reverse engineering at 32K context

  1. 73Great quality (A) · 73 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. Ornith 1.5 9B Q4_K_M · 1.8 tok/s · Very slow
  2. 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.5 9B Q4_K_M · 1.8 tok/s · Very slow
  3. 74Great quality (A) · 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. SenseNova Loop 9B Preview Q4_K_M · 1.8 tok/s · Very slow (tight fit)
  4. 71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. Ornith 1.0 9B Q4_K_M · 1.8 tok/s · Very slow
  5. 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Bolt E4B Q5_K_M · 3.2 tok/s · Slow

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?