The AI Bench
48 GB · 768 GB/s

What can the NVIDIA RTX A6000 (used) run?

48 GB on one card. Slower than new cards, but fits 70B models.

Best models for each task

Our top three for this machine. Chat and image tasks assume an 8K-token conversation; coding, documents and agents assume 32K.

Chat & writing20 of 25 run

  1. Qwen3.8 27BRuns greatQ8_0 · needs 31.6 GB at 8K · 10–16tok/sReadable
  2. Gemma 4 31BRuns wellQ8_0 · needs 35.7 GB at 8K · 9–14tok/sReadable
  3. Muse Glimmer 30BRuns greatQ8_0 · needs 32.1 GB at 8K · 10–16tok/sReadable

Coding21 of 26 run

  1. Qwen3.8 27BRuns greatQ4_K_M · needs 21.5 GB at 32K · 15–25tok/sSmooth
  2. Muse Glimmer 30BRuns greatQ4_K_M · needs 20.2 GB at 32K · 17–25tok/sSmooth
  3. Qwen3.6 35B-A3BRuns wellQ8_0 · needs 40.4 GB at 32K · 40–85tok/sFast

Long documents15 of 20 run

  1. Qwen3.8 27BRuns greatQ4_K_M · needs 21.5 GB at 32K · 15–25tok/sSmooth
  2. Qwen3.6 35B-A3BRuns wellQ8_0 · needs 40.4 GB at 32K · 40–85tok/sFast
  3. Gemma 4 26B-A4BRuns greatQ8_0 · needs 29.9 GB at 32K · 35–70tok/sFast

Agents & tools21 of 26 run

  1. Qwen3.8 27BRuns greatQ4_K_M · needs 21.5 GB at 32K · 15–25tok/sSmooth
  2. Qwen3.6 35B-A3BRuns wellQ8_0 · needs 40.4 GB at 32K · 40–85tok/sFast
  3. Ornith 1.5 35B-A3BRuns wellQ8_0 · needs 39.5 GB at 32K · 40–85tok/sFast

Images in9 of 13 run

  1. Qwen3.8 27BRuns greatQ8_0 · needs 31.6 GB at 8K · 10–16tok/sReadable
  2. Qwen3.6 35B-A3BRuns wellQ8_0 · needs 39.9 GB at 8K · 50–95tok/sFast
  3. Muse Glimmer 30BRuns greatQ8_0 · needs 32.1 GB at 8K · 10–16tok/sReadable

Make images6 of 6 run

  1. Stable Diffusion XL Base 1.0Runs greatFP16 · needs 7.9 GB
  2. FLUX.2 [klein] 4BRuns greatBF16 · needs 14.0 GB
  3. Z-Image-TurboRuns greatBF16 · needs 17.0 GB

Speak text7 of 7 run

  1. Qwen3-TTS 1.7BRuns greatBF16 · needs 5.5 GB
  2. Kokoro-82MRuns greatFP32 · needs 1.4 GB
  3. Chatterbox Multilingual V3Runs greatFull · needs 4.2 GB

Transcribe7 of 7 run

  1. Voxtral Mini 4B RealtimeRuns greatBF16 · needs 17.0 GB
  2. Whisper large-v3-turbo (via faster-whisper)Runs greatFP16 · needs 2.6 GB
  3. Nemotron 3.5 ASR Streaming 0.6BRuns greatFP32 · needs 3.5 GB

Biggest model it runs well

Kimi Linear 48B-A3B (Q4_K_M, 30.1 GB file) at about 75–140tok/sFast.

Should you buy it?

Buy it if

You need 48 GB in one slot (workstation or server).

Skip it if

Two used RTX 3090s give 48 GB for about half the price.

Watch out for

  • Used cards are sold as-is; test memory on arrival.