Cheapest box to run Mistral-Large-2411 at BF16
Mistral’s Mistral-Large-2411 is 123B parameters, dense. 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.
Mistral-Large-2411 at BF16 needs 250GB.
Cheapest Config that fits: Mac Studio, M5 Ultra, 512GB. Fits, 254GB headroom. Est. 3.2 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−237GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−229GB headroom
- Mac mini, M6, 32GB$1,499Won't fit−221GB headroom
- Mac mini, M5 Pro, 24GB$1,699Won't fit−229GB headroom
- Mac mini, M5 Pro, 48GB$2,499Won't fit−206GB headroom
- Mac Studio, M5 Max, 36GB$2,499Won't fit−217GB headroom
- Mac mini, M5 Pro, 64GB$2,899Won't fit−191GB headroom
- GeForce RTX 5090, 32GB$2,999.99+hostWon't fit−219GB headroom
- Mac Studio, M5 Max, 48GB$3,099Won't fit−206GB headroom
- Mac Studio, M5 Max, 64GB$3,499Won't fit−191GB headroom
- Won't fit−130GB headroom
- Won't fit−130GB headroom
- Won't fit−130GB headroom
- Won't fit−162GB headroom
- Won't fit−2GB headroom
- Fits254GB headroomest. 3.2 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −237GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −229GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Won't fit | −221GB | — |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −229GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Won't fit | −206GB | — |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Won't fit | −217GB | — |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Won't fit | −191GB | — |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Won't fit | −219GB | — |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Won't fit | −206GB | — |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Won't fit | −191GB | — |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Won't fit | −130GB | — |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Won't fit | −130GB | — |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Won't fit | −130GB | — |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Won't fit | −162GB | — |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Won't fit | −2GB | — |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 254GB | 3.2 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 = 123B × 2.000 B/param (16 bits) + 0.360 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.65 ÷ (123B active × 2.000 B/param) = 246.0 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.