aliteq.

Type to search every guide, tool and GPU page.

How much VRAM gpt-oss 120B (MoE) needs, and what it costs to run

Some links on this page are referral links: we may earn a commission, at no cost to you.

Trying to decide what to buy? The best GPU for gpt-oss 120B (MoE) ↓ · Got a card already? See what your PC can run →

116.8B paramsMoE · 128 experts, 4 active128k max contextapache-2.0

Cheapest way to rent it right now

$0.24/hr

NVIDIA DGX Spark (GB10 Grace Blackwell) · MXFP4 · on-demand · ≈ $58/mo at 8h/day

Live price captured Thu, 19 Nov 2026 17:20:08 GMT.It never changes which card is cheapest.

Just want to use it? Per-token API

$0.05in·$0.25out

per 1M tokens · OpenAI GPT OSS 120B on Novita AI · read Oct 29, 2026

No GPU to manage: you pay per token, which suits occasional use. Your prompts run on Novita's servers, not your own machine.

Price as listed on Novita's pricing page on Oct 29, 2026; it may have changed.

Rent a GPU or pay per token?

A per-token API charges for every token, so it is cheapest when you use gpt-oss 120B (MoE) a little. A rented GPU charges by the hour whether it is busy or idle, so it only pays off once you push enough tokens through it. Set your daily volume and see where the line falls for this model, using the cheapest live rental that fits and that we have a memory-bandwidth figure for (DGX Spark (GB10 Grace Blackwell), MXFP4, $0.240/hr on-demand) and Novita AI's published rate.

Tokens per day (input + output)
Input : output tokens
  • Pay per token · Novita AI$3.00/mo

    $0.10/day · $0.10 per 1M tokens blended at 3:1

  • Rent DGX Spark (GB10 Grace Blackwell) · Vast.ai$9.05/mo

    $0.30/day · 1.3 GPU-hours needed

At 1M tokens/day the API is cheaper. The API is cheaper at every volume: at our estimated speed, renting the DGX Spark (GB10 Grace Blackwell) works out to about $0.30 per 1M tokens against the API's $0.10.

At a glance · 3:1 input:output

Tokens/dayAPI /moRent /mo
100K$0.30$7.21
1M$3.00$9.05
10M$30.00$90.54
100M$300$905

Cheaper option in green. 30-day month. Break-even: none, the API stays cheaper.

How we work it out. API cost = input tokens × $0.05 + output tokens × $0.25 per 1M. Rent cost = the hourly price × the GPU time needed, with at least 1 hour billed per day (spin-up, model load and idle time cost money too). GPU time assumes the DGX Spark (GB10 Grace Blackwell) writes ~72 output tokens/sec for gpt-oss 120B (MoE) at MXFP4 (our estimate: its 273 GB/s memory bandwidth from the spec sheet, not yet confirmed by a second source ÷ the bytes of weights read per token, × 75% efficiency, one request at a time) and reads prompt tokens 10× faster than it writes. Above 24 GPU-hours a day you would need more than one card.

These are estimates: the tokens/sec figure is our model, not a benchmark, and batching many requests (vLLM, SGLang) can raise a rented card's throughput a lot, which moves the break-even lower. Rental price captured Sat, 21 Nov 2026 08:20:09 GMT. API price as listed by Novita AI on Oct 29, 2026; API prices change.

Memory needed, by quantisation

At 16k context. Weights + KV cache + ~0.8 GB overhead. gpt-oss 120B (MoE) ships in MXFP4, so its own release leads and no bigger copy is listed: it would only store the same numbers in more bytes. Memory is in binary GB (1,024³ bytes), the unit GPU memory is sold in; Hugging Face and Ollama list downloads in decimal GB, about 7% more.

QuantWeightsKV cacheTotalQuality
MXFP4official60.8 GB1.1 GB62.7 GBThe model's own release (MXFP4, a 65.2 GB download): the best quality there is. A bigger copy only stores the same numbers in more bytes. Its 4-bit GGUF builds keep the 4-bit experts as they are, so they come out about this size too.

Cards that can run it

Best quantisation each card fits at 16k context, with live rental price where we track one.

GPUVRAMBest fitUses~tok/sRent from
Apple M3 Ultra512 GBMXFP463 GB——
Apple M5 Ultra512 GBMXFP463 GB——
Apple M2 Ultra192 GBMXFP463 GB——
Apple M5 Max (18-core CPU / 40-core GPU)128 GBMXFP463 GB——
NVIDIA DGX Spark (GB10 Grace Blackwell)128 GBMXFP463 GB—$0.240/hrVast.ai
Apple M4 Max (16-core CPU / 40-core GPU)128 GBMXFP463 GB——
AMD Ryzen AI Max+ 395 (Strix Halo, 128 GB)128 GBMXFP463 GB——
NVIDIA RTX PRO 6000 Blackwell Workstation Edition96 GBMXFP463 GB—$0.533/hrVast.ai
Apple M2 Max96 GBMXFP463 GB——
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition96 GBMXFP463 GB—$0.333/hrVast.ai
Apple M4 Pro64 GBMXFP4tight63 GB——

The best GPU for gpt-oss 120B (MoE)

Ranked from live rental prices and computed VRAM fit, not opinion. Every card is judged at its own MXFP4 release with a 8K context, so the comparison is fair; a card that could only run a crushed version isn't listed.

Best value

NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition

Best throughput per dollar among cards that fit.

Cheapest that runs it

NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition

Lowest hourly rental that fits it — $0.333/hr, MXFP4.

No compromise

—

Most headroom and highest throughput.

Every card that runs it, ranked

Sorted by tokens/sec per rental dollar. Throughput is a bandwidth-derived estimate.

GPUVRAMVRAM used~tok/sCheapest rentaltok/s per $
NVIDIA RTX PRO 6000 Blackwell Workstation Edition96 GB62 GB65%—$0.533—
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition96 GB62 GB65%—$0.333—

The value column ranks renting, because that's what we can price precisely (prices captured Thu, 19 Nov 2026 17:20:08 UTC). Buying a card? The fit and throughput columns are what you need; we don't publish a purchase-price ranking because we don't hold verified street prices. Tokens/sec is an estimate, not a benchmark. Got a card already? See everything your PC can run.

gpt-oss-120b

How these numbers are produced

  • Model shape comes from the model's own config.json — 36 layers, 64 attention heads, 8 KV heads.
  • Weight size is parameters × effective bits-per-weight. Those constants are checked against real published quantised file sizes — median error 0.7% across 40 measurements.
  • KV cache is 2 × layers × KV-heads × head-dim × context × 2 bytes. Using KV-heads rather than attention heads is what makes this correct for grouped-query attention; treating a GQA model as multi-head overstates the cache by up to 8×.
  • Tokens/sec is an estimate, not a benchmark. Generation is memory-bandwidth-bound, so this is bandwidth ÷ weight-bytes derated to 75%. Real throughput depends on your runtime, batch size and kernels. We don't run our own hardware tests — see our editorial policy.
  • This is a mixture-of-experts model. Generation reads only the experts active for each token (4 of 128), so throughput is scored on ~5.1B active parameters, not the full 117B. Memory is the opposite: every expert must still be resident in VRAM, so the figures above use the full weight set.Active = 5.1B, vendor figure: OpenAI model card (huggingface.co/openai/gpt-oss-120b, read 29 Oct 2026): "117B parameters with 5.1B active parameters". params = HF safetensors total. Also set in workers/prices/src/models.ts seeds so the daily ingest keeps it.
  • Some memory-bandwidth figures are not yet independently verified.Where that's the case we show no tokens/sec at all rather than a number we can't stand behind.
  • Rental prices are pulled hourly from provider APIs (last updated Thu, 19 Nov 2026 17:20:08 GMT). Spot/interruptible pricing can change or vanish without notice.

Try it interactively

Drag the context slider and watch the KV cache fill the card

Running gpt-oss 120B (MoE): common questions

How much VRAM do you need to run gpt-oss 120B (MoE)?
In its own MXFP4 release and at 8k context, gpt-oss 120B (MoE) needs about 62.1 GB — 60.8 GB of weights plus 0.6 GB of KV cache and ~0.8 GB overhead. So you want a card with at least that much free memory; the KV cache grows if you use longer context.
Can an RTX 4090 run gpt-oss 120B (MoE)?
Not in its own MXFP4 release — gpt-oss 120B (MoE) needs more than the card's 24 GB at that quality and 8k context. Its MXFP4 release is already the small build (its 4-bit GGUFs are about the same size), so you'd need a bigger card, two cards, or to spill part of it into system RAM, which is much slower.
Can an RTX 3090 run gpt-oss 120B (MoE)?
Not in its own MXFP4 release — gpt-oss 120B (MoE) needs more than the card's 24 GB at that quality and 8k context. Its MXFP4 release is already the small build (its 4-bit GGUFs are about the same size), so you'd need a bigger card, two cards, or to spill part of it into system RAM, which is much slower.
Can an RTX 4060 Ti 16GB run gpt-oss 120B (MoE)?
Not in its own MXFP4 release — gpt-oss 120B (MoE) needs more than the card's 16 GB at that quality and 8k context. Its MXFP4 release is already the small build (its 4-bit GGUFs are about the same size), so you'd need a bigger card, two cards, or to spill part of it into system RAM, which is much slower.
What's the cheapest way to run gpt-oss 120B (MoE)?
Among cards that fit it in its own MXFP4 release, the cheapest to rent right now is the NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition at about $0.333/hour. Whether renting or buying is cheaper overall depends on how many hours a day you'll actually use it.

Worth reading before you buy

Other models