Cheapest box to run gpt-oss-120b at BF16
OpenAI’s gpt-oss-120b is 117B parameters, 5.1B of them read per token. The number below is computed from that and each vendor’s published specs, not asserted. Switch the quant or the context and it recomputes in place.
gpt-oss-120b at BF16 needs 236GB.
Cheapest Config that fits: Mac Studio, M5 Ultra, 512GB. Fits, 268GB headroom. Est. 58.8 tok/s.
No price yet: Ships late October 2026. Apple has not published a price.
Collecting price history — 2 of 90 days. No timing verdict yet.
Offers — Mac Studio, M5 Ultra, 512GB, 1TB
Ships late October 2026. Apple has not published a price.
- Mac mini, M6, 16GB$1,099Won't fit−223GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−215GB headroom
- Mac mini, M6, 32GB$1,499Won't fit−207GB headroom
- Mac mini, M5 Pro, 24GB$1,699Won't fit−215GB headroom
- Mac mini, M5 Pro, 48GB$2,499Won't fit−192GB headroom
- Mac Studio, M5 Max, 36GB$2,499Won't fit−203GB headroom
- Mac mini, M5 Pro, 64GB$2,899Won't fit−177GB headroom
- GeForce RTX 5090, 32GB$2,999.99+hostWon't fit−205GB headroom
- Mac Studio, M5 Max, 48GB$3,099Won't fit−192GB headroom
- Mac Studio, M5 Max, 64GB$3,499Won't fit−177GB headroom
- Won't fit−115GB headroom
- Won't fit−115GB headroom
- Won't fit−116GB headroom
- Won't fit−147GB headroom
- Tight12GB headroomest. 58.8 tok/s
- Fits268GB headroomest. 58.8 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −223GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −215GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Won't fit | −207GB | — |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −215GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Won't fit | −192GB | — |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Won't fit | −203GB | — |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Won't fit | −177GB | — |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Won't fit | −205GB | — |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Won't fit | −192GB | — |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Won't fit | −177GB | — |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Won't fit | −115GB | — |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Won't fit | −115GB | — |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Won't fit | −116GB | — |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Won't fit | −147GB | — |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Tight | 12GB | 58.8 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 268GB | 58.8 tok/s |
GeForce RTX 5090 is a card, not a computer: every price shown for it is the cheapest listing we found plus $1,000 for the minimum host machine.
memory = 117B × 2.000 B/param (16 bits) + 0.074 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (5.1B active × 2.000 B/param) = 10.2 GB read per token
usable = memory − operating system reserve (8% on a Mac, 6% on a DGX Spark, 4% on a card)
Bandwidth is the vendor’s published figure. Every speed is an estimate and says so. Methodology.
What else Mac Studio M5 Ultra 512GB runs
- DeepSeek-V3Q5_K_M, est. 22.9 tok/s
- Llama-4-Maverick-17B-128EQ8_0, est. 33.2 tok/s
- Qwen3-235B-A22BBF16, est. 13.6 tok/s
- Mistral-Large-2411BF16, est. 3.2 tok/s
- gpt-oss-120bBF16, est. 58.8 tok/s
- Llama-4-Scout-17B-16EBF16, est. 17.6 tok/s
Best quant that still fits at 8K context, largest model first. Full list on the Config page.
Bandwidth 1.2TB/s, usable memory 504GB of 512GB.