← Every GPUAMD · 32 or 16 or 64 GB system memory · 120 GB/s

What can the Radeon 890M run?

With 32 GB, 6 of 12 popular open models fit. The best pick is gpt-oss 20B at about 21 tok/s, about reading speed.

ollama run gpt-oss:20b

First time? Install Ollama, then paste that into a terminal. Is this your machine? Check it on the homepage.

16 GB system memory · about 10 GB for models

Every model the Radeon 890M 16 GB can run

3 of 12 fit. Best pick: Qwen3 4B.

  • Qwen3 4BQuick everyday chat and summaries
    ~26 tok/s
    Usableqwen3:4b
  • Qwen3 8BA solid all-rounder for chat and writing
    ~13 tok/s
    Usableqwen3:8b
  • Gemma 3 12BReads images as well as text
    ~8 tok/s
    Slowgemma3:12b
  • Qwen3 14BNoticeably smarter at reasoning and code
    needs 11 GB
    Too bigqwen3:14b
  • gpt-oss 20BOpenAI's open model. Fast, strong reasoning
    needs 16 GB
    Too biggpt-oss:20b
  • Gemma 3 27BGoogle's best open model for its size
    needs 19 GB
    Too biggemma3:27b
  • Qwen3 30B-A3B30B smarts at small-model speed
    needs 21 GB
    Too bigqwen3:30b
  • Qwen3 32BStrong coding and long answers
    needs 22 GB
    Too bigqwen3:32b
  • Llama 3.3 70BClose to the big cloud models
    needs 45 GB
    Too bigllama3.3:70b
  • gpt-oss 120BOpenAI's largest open model
    needs 67 GB
    Too biggpt-oss:120b
  • Qwen3 235B-A22BAlibaba's flagship open model, for multi-GPU rigs
    needs 144 GB
    Too bigqwen3:235b
  • DeepSeek R1 671BThe full DeepSeek R1, for the biggest rigs
    needs 406 GB
    Too bigdeepseek-r1:671b
32 GB system memory · about 19 GB for models

Every model the Radeon 890M 32 GB can run

6 of 12 fit. Best pick: gpt-oss 20B.

  • Qwen3 4BQuick everyday chat and summaries
    ~26 tok/s
    Usableqwen3:4b
  • Qwen3 8BA solid all-rounder for chat and writing
    ~13 tok/s
    Usableqwen3:8b
  • Gemma 3 12BReads images as well as text
    ~8 tok/s
    Slowgemma3:12b
  • Qwen3 14BNoticeably smarter at reasoning and code
    ~7 tok/s
    Slowqwen3:14b
  • gpt-oss 20BOpenAI's open model. Fast, strong reasoning
    ~21 tok/s
    Usablegpt-oss:20b
  • Gemma 3 27BGoogle's best open model for its size
    ~4 tok/s
    Slowgemma3:27b
  • Qwen3 30B-A3B30B smarts at small-model speed
    needs 21 GB
    Too bigqwen3:30b
  • Qwen3 32BStrong coding and long answers
    needs 22 GB
    Too bigqwen3:32b
  • Llama 3.3 70BClose to the big cloud models
    needs 45 GB
    Too bigllama3.3:70b
  • gpt-oss 120BOpenAI's largest open model
    needs 67 GB
    Too biggpt-oss:120b
  • Qwen3 235B-A22BAlibaba's flagship open model, for multi-GPU rigs
    needs 144 GB
    Too bigqwen3:235b
  • DeepSeek R1 671BThe full DeepSeek R1, for the biggest rigs
    needs 406 GB
    Too bigdeepseek-r1:671b
64 GB system memory · about 38 GB for models

Every model the Radeon 890M 64 GB can run

8 of 12 fit. Best pick: Qwen3 30B-A3B.

  • Qwen3 4BQuick everyday chat and summaries
    ~26 tok/s
    Usableqwen3:4b
  • Qwen3 8BA solid all-rounder for chat and writing
    ~13 tok/s
    Usableqwen3:8b
  • Gemma 3 12BReads images as well as text
    ~8 tok/s
    Slowgemma3:12b
  • Qwen3 14BNoticeably smarter at reasoning and code
    ~7 tok/s
    Slowqwen3:14b
  • gpt-oss 20BOpenAI's open model. Fast, strong reasoning
    ~21 tok/s
    Usablegpt-oss:20b
  • Gemma 3 27BGoogle's best open model for its size
    ~4 tok/s
    Slowgemma3:27b
  • Qwen3 30B-A3B30B smarts at small-model speed
    ~21 tok/s
    Usableqwen3:30b
  • Qwen3 32BStrong coding and long answers
    ~3 tok/s
    Slowqwen3:32b
  • Llama 3.3 70BClose to the big cloud models
    needs 45 GB
    Too bigllama3.3:70b
  • gpt-oss 120BOpenAI's largest open model
    needs 67 GB
    Too biggpt-oss:120b
  • Qwen3 235B-A22BAlibaba's flagship open model, for multi-GPU rigs
    needs 144 GB
    Too bigqwen3:235b
  • DeepSeek R1 671BThe full DeepSeek R1, for the biggest rigs
    needs 406 GB
    Too bigdeepseek-r1:671b
Common questions

Can the Radeon 890M run it?

Can the Radeon 890M run Qwen3 30B-A3B?

Yes, with 64 GB or more. Qwen3 30B-A3B needs about 21 GB and the Radeon 890M with 64 GB of system memory leaves about 38 GB for models. Expect about 21 tok/s, about reading speed. The 16 and 32 GB versions don't have room.

Can the Radeon 890M run Qwen3 32B?

Yes, with 64 GB or more. Qwen3 32B needs about 22 GB and the Radeon 890M with 64 GB of system memory leaves about 38 GB for models. Expect about 3 tok/s, slow, but it works. The 16 and 32 GB versions don't have room.

Can the Radeon 890M run Llama 3.3 70B?

No. Llama 3.3 70B needs about 45 GB and the Radeon 890M has at most 64 GB of system memory, of which about 38 GB is usable for models. The most capable model it runs well is Qwen3 30B-A3B, at about 21 tok/s.

How we estimate

Generating a token reads the model's active weights from memory once, so speed is roughly memory bandwidth (120 GB/s) divided by the active weight size, with efficiency calibrated on our own rack. Sizes are Ollama's default 4-bit downloads plus about 1.2 GB for context. Built-in graphics share system memory, so we count about 60% of it. The easiest start is a small model in your browser in the Studio. Real results are usually within about a third. Measured one yourself? Send us the number.

Want something bigger?

The Used workstation + 2× Tesla P100 runs Qwen3 32B at about 20 tok/s, from $750 used.

See the builds

Compare

Similar cards

Every GPU we track