Best local LLMs for the NVIDIA H20

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. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Kolibri 1 (78B-A3.5B) Q6_K · 216 tok/s · Runs great
  2. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Super VL 120B-A12B Q4_K_M · 172 tok/s · Runs great
  3. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. GLM-4.7 Flash (30B-A3B) F16 · 143 tok/s · Runs great
  4. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Raon-1 mini Preview (30B-A4B) F16 · 123 tok/s · Runs great
  5. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Ornith 1.5 122B-A10B Q5_K_M · 168 tok/s · Runs well (tight fit)

Coding at 32K context

  1. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Kolibri 1 (78B-A3.5B) Q6_K · 216 tok/s · Runs great
  2. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Super VL 120B-A12B Q4_K_M · 172 tok/s · Runs great
  3. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. GLM-4.7 Flash (30B-A3B) F16 · 143 tok/s · Runs great
  4. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Raon-1 mini Preview (30B-A4B) F16 · 123 tok/s · Runs great
  5. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Bolt 74B-A9B Q5_K_M · 175 tok/s · Runs great

Reasoning at 32K context

  1. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Kolibri 1 (78B-A3.5B) Q6_K · 216 tok/s · Runs great
  2. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Super VL 120B-A12B Q4_K_M · 172 tok/s · Runs great
  3. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Raon-1 mini Preview (30B-A4B) F16 · 123 tok/s · Runs great
  4. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Fara 1.5 27B F16 · 42.1 tok/s · Runs great
  5. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Bolt 74B-A9B Q5_K_M · 175 tok/s · Runs great

Documents (RAG) at 32K context

  1. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Kolibri 1 (78B-A3.5B) Q6_K · 216 tok/s · Runs great
  2. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Super VL 120B-A12B Q4_K_M · 172 tok/s · Runs great
  3. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. GLM-4.7 Flash (30B-A3B) F16 · 143 tok/s · Runs great
  4. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Fara 1.5 27B F16 · 42.1 tok/s · Runs great
  5. 89Excellent quality (A+) · 89 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.5 122B-A10B Q5_K_M · 168 tok/s · Runs well (tight fit)

Agents & tools at 64K context

  1. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. Kolibri 1 (78B-A3.5B) Q6_K · 216 tok/s · Runs great
  2. 90Excellent quality (A+) · 90 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Nemotron 3.5 Super VL 120B-A12B Q5_K_M · 155 tok/s · Runs well (tight fit)
  3. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Fara 1.5 27B F16 · 42.1 tok/s · Runs great
  4. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. GLM-4.7 Flash (30B-A3B) F16 · 143 tok/s · Runs great
  5. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. Raon-1 mini Preview (30B-A4B) F16 · 123 tok/s · Runs great

Reverse engineering at 32K context

  1. 88Excellent quality (A+) · 88 of 100: how capable the model is at F16, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. GLM-4.7 Flash (30B-A3B) F16 · 143 tok/s · Runs great
  2. 95Excellent quality (A+) · 95 of 100: how capable the model is at F16, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.8 27B F16 · 41.5 tok/s · Runs great
  3. 84Great quality (A) · 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. gpt-oss 120B MXFP4 · 231 tok/s · Runs great
  4. 93Excellent quality (A+) · 93 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.5 35B-A3B Q8_0 · 200 tok/s · Runs great
  5. 94Excellent quality (A+) · 94 of 100: how capable the model is at F16, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. Qwen3.6 27B F16 · 41.5 tok/s · Runs great

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?