Can GPT-OSS 120B run on AMD Instinct MI250 128GB?

YES — Runs Great

S94Excellent

Estimated from fit model

GPT-OSS 120B needs ~90.0 GB VRAM. AMD Instinct MI250 128GB has 128.0 GB. With Q4_K_M quantization, expect ~33 tok/s.

Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Balanced

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 90.0 GB, 33.2 tok/s, Runs well

90.0 GB required128.0 GB available

70% VRAM used

Fit status

Runs well

Decode

33.2 tok/s

TTFT

5839 ms

Safe context

131K

Memory

90.0 GB / 128.0 GB

Memory breakdown

Weights71.4 GB

KV Cache4.9 GB

Runtime0.9 GB

Headroom12.8 GB

See how fast it feels

See how fast it feelsGPT-OSS 120B on AMD Instinct MI250 128GB
1st promptCold start — includes initialization
>Write a Python function to sort a list
Here's a Python function that sorts a list using quicksort:

def quicksort(arr):
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left   = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right  = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

# Example usage
numbers = [3, 6, 8, 10, 1, 2, 1]
print(quicksort(numbers))  # [1, 1, 2, 3, 6, 8, 10]

The algorithm picks a pivot, partitions the list into smaller, equal, and larger elements, then recursively sorts each partition and concatenates the results.
2nd promptWarm — model ready, no init overhead
>Now add type hints and a docstring
def quicksort(arr: list[int]) -> list[int]:
    """Sort a list using the quicksort algorithm.

    Args:
        arr: List of integers to sort.

    Returns:
        A new sorted list.
    """
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left   = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right  = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

Added type hints for the input and return type, plus a Google-style docstring.
Estimated: 33.2 tok/s decode · 5.8s TTFT (warm) · 83 tok/s prefill

What limits this setup

This setup is broadly balanced for this model.

No major red flags

This recommendation has enough memory headroom and acceptable estimated speed for the selected workload.

Best improvement path

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	S	Runs well	33.2 tok/s	3185 ms	131K
Coding	S	Runs well	33.2 tok/s	5839 ms	131K
Agentic Coding	S	Runs well	33.2 tok/s	8493 ms	131K
Reasoning	S	Runs well	33.2 tok/s	6901 ms	131K
RAG	S	Runs well	33.2 tok/s	10616 ms	131K

Inference speed

GPT-OSS 120B inference speed — tokens per second by GPU & Mac

Estimated decode speed (tokens/sec) for GPT-OSS 120B at Q4_K_M across popular GPUs and Apple Silicon, including multi-GPU rigs, using the fastest local runtime per device. Fastest is MacBook Pro M4 Max 128GB at ~9 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.

GPU / Mac	Memory	Quant	Speed (tok/s)	Fits?
MacBook Pro M4 Max 128GB	128 GB	Q4_K_M	9.2	Offloads
Mac Studio M3 Ultra 256GB	256 GB	Q4_K_M	8.5	Fits
Mac Studio M2 Ultra 128GB	128 GB	Q4_K_M	7.1	Offloads
Mac Studio M1 Ultra 128GB	128 GB	Q4_K_M	6.7	Offloads
2× RX 7900 XTX 24GB	48 GB	Q4_K_M	4.4	Too big
MacBook Pro M4 Max 64GB	64 GB	Q4_K_M	4.3	Too big
2× RTX 4090 24GB	48 GB	Q4_K_M	3.1	Too big
MacBook Pro M4 Pro 48GB	48 GB	Q4_K_M	2.6	Too big
2× RTX 3090 24GB	48 GB	Q4_K_M	2.6	Too big
4× RTX 3060 12GB	48 GB	Q4_K_M	2.3	Too big
RTX 5090 32GB	32 GB	Q4_K_M	2.0	Too big
RTX 4090 24GB	24 GB	Q4_K_M	2.0	Too big
RTX 4080 Super 16GB	16 GB	Q4_K_M	2.0	Too big
RTX 3090 24GB	24 GB	Q4_K_M	2.0	Too big
RTX 4070 12GB	12 GB	Q4_K_M	2.0	Too big
RTX 3060 12GB	12 GB	Q4_K_M	2.0	Too big
RTX 4060 8GB	8 GB	Q4_K_M	2.0	Too big
RX 7900 XTX 24GB	24 GB	Q4_K_M	2.0	Too big
MacBook Pro M3 Max 64GB	64 GB	Q4_K_M	2.0	Too big
MacBook Pro M1 Max 64GB	64 GB	Q4_K_M	2.0	Too big

Estimates for single-stream decoding at Q4_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.

Quantization options

How GPT-OSS 120B (117B params) fits at each quantization level on AMD Instinct MI250 128GB (128.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	45.6 GB	Low	A84
Q3_K_S	3	57.3 GB	Low	S86
NVFP4	4	65.5 GB	Medium	S88
Q4_K_M	4	71.4 GB	Medium	S88
Q5_K_M	5	84.2 GB	High	S88
Q6_KBest for your GPU	6	95.9 GB	High	S88
Q8_0	8	125.2 GB	Very High	F0
F16	16	239.8 GB	Maximum	F0

Get started

Copy-paste commands to run GPT-OSS 120B on your machine.

Run

ollama run gpt-oss:120b

Your hardware

More models your AMD Instinct MI250 128GB can run

Model	Params	Grade	Decode
Devstral 2 123B Instruct	123B	S	31.5 tok/s
Qwen 3.5 122B A10B	122B	S	87.5 tok/s
Mistral Small 4 119B	119B	S	94.8 tok/s

Frequently asked questions

Can AMD Instinct MI250 128GB run GPT-OSS 120B?

Yes, AMD Instinct MI250 128GB can run GPT-OSS 120B with a S grade (Runs well). Expected decode speed: 33.2 tok/s.

How much VRAM does GPT-OSS 120B need?

GPT-OSS 120B (117B parameters) requires approximately 90.0 GB of memory with Q4_K_M quantization.

What is the best quantization for GPT-OSS 120B?

The recommended quantization for GPT-OSS 120B is Q4_K_M, which balances quality and memory efficiency.

What speed will GPT-OSS 120B run at on AMD Instinct MI250 128GB?

On AMD Instinct MI250 128GB, GPT-OSS 120B achieves approximately 33.2 tokens per second decode speed with a time-to-first-token of 5839ms using Q4_K_M quantization.

Can AMD Instinct MI250 128GB run GPT-OSS 120B for coding?

For coding workloads, GPT-OSS 120B on AMD Instinct MI250 128GB receives a S grade with 33.2 tok/s and 131K context.

What context window can GPT-OSS 120B use on AMD Instinct MI250 128GB?

On AMD Instinct MI250 128GB, GPT-OSS 120B can safely use up to 131K tokens of context. The model's official context limit is 131K, but available memory constrains the safe maximum.

See all results for AMD Instinct MI250 128GB See all hardware for GPT-OSS 120B

Embed this result▼

Paste this snippet into any page to show a live fit card.

<iframe src="https://willitrunai.com/embed/gpt-oss-120b-on-instinct-mi250-128gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview:

Workload

Grade

Fit

Decode

TTFT

Context

Chat

Runs well

33.2 tok/s

3185 ms

131K

Coding

Runs well

33.2 tok/s

5839 ms

131K

Agentic Coding

Runs well

33.2 tok/s

8493 ms

131K

Reasoning

Runs well

33.2 tok/s

6901 ms

131K

RAG

Runs well

33.2 tok/s

10616 ms

131K

GPT-OSS 120B inference speed — tokens per second by GPU & Mac

GPU / Mac

Memory

Quant

Speed (tok/s)

Fits?

MacBook Pro M4 Max 128GB

128 GB

Q4_K_M

9.2

Offloads

Mac Studio M3 Ultra 256GB

256 GB

Q4_K_M

8.5

Fits

Mac Studio M2 Ultra 128GB

128 GB

Q4_K_M

7.1

Offloads

Mac Studio M1 Ultra 128GB

128 GB

Q4_K_M

6.7

Offloads

2× RX 7900 XTX 24GB

48 GB

Q4_K_M

4.4

Too big

MacBook Pro M4 Max 64GB

64 GB

Q4_K_M

4.3

Too big

2× RTX 4090 24GB

48 GB

Q4_K_M

3.1

Too big

MacBook Pro M4 Pro 48GB

48 GB

Q4_K_M

2.6

Too big

2× RTX 3090 24GB

48 GB

Q4_K_M

2.6

Too big

4× RTX 3060 12GB

48 GB

Q4_K_M

2.3

Too big

RTX 5090 32GB

32 GB

Q4_K_M

2.0

Too big

RTX 4090 24GB

24 GB

Q4_K_M

2.0

Too big

RTX 4080 Super 16GB

16 GB

Q4_K_M

2.0

Too big

RTX 3090 24GB

24 GB

Q4_K_M

2.0

Too big

RTX 4070 12GB

12 GB

Q4_K_M

2.0

Too big

RTX 3060 12GB

12 GB

Q4_K_M

2.0

Too big

RTX 4060 8GB

8 GB

Q4_K_M

2.0

Too big

RX 7900 XTX 24GB

24 GB

Q4_K_M

2.0

Too big

MacBook Pro M3 Max 64GB

64 GB

Q4_K_M

2.0

Too big

MacBook Pro M1 Max 64GB

64 GB

Q4_K_M

2.0

Too big

Quantization options

How GPT-OSS 120B (117B params) fits at each quantization level on AMD Instinct MI250 128GB (128.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	45.6 GB	Low	A84
Q3_K_S	3	57.3 GB	Low	S86
NVFP4	4	65.5 GB	Medium	S88
Q4_K_M	4	71.4 GB	Medium	S88
Q5_K_M	5	84.2 GB	High	S88
Q6_KBest for your GPU	6	95.9 GB	High	S88
Q8_0	8	125.2 GB	Very High	F0
F16	16	239.8 GB	Maximum	F0

Model

Params

Grade

Decode

Capabilities

Devstral 2 123B Instruct

123B

31.5 tok/s

Qwen 3.5 122B A10B

122B

87.5 tok/s

Mistral Small 4 119B

119B

94.8 tok/s

Frequently asked questions

Can AMD Instinct MI250 128GB run GPT-OSS 120B?

Yes, AMD Instinct MI250 128GB can run GPT-OSS 120B with a S grade (Runs well). Expected decode speed: 33.2 tok/s.

How much VRAM does GPT-OSS 120B need?

GPT-OSS 120B (117B parameters) requires approximately 90.0 GB of memory with Q4_K_M quantization.

What is the best quantization for GPT-OSS 120B?

The recommended quantization for GPT-OSS 120B is Q4_K_M, which balances quality and memory efficiency.

What speed will GPT-OSS 120B run at on AMD Instinct MI250 128GB?

On AMD Instinct MI250 128GB, GPT-OSS 120B achieves approximately 33.2 tokens per second decode speed with a time-to-first-token of 5839ms using Q4_K_M quantization.

Can AMD Instinct MI250 128GB run GPT-OSS 120B for coding?

For coding workloads, GPT-OSS 120B on AMD Instinct MI250 128GB receives a S grade with 33.2 tok/s and 131K context.

What context window can GPT-OSS 120B use on AMD Instinct MI250 128GB?

On AMD Instinct MI250 128GB, GPT-OSS 120B can safely use up to 131K tokens of context. The model's official context limit is 131K, but available memory constrains the safe maximum.