Xiaomi · Chat and reasoning · 1T parameters · 42B active · mit
MiMo V2.6 Pro RL
Needs 689 GB at INT4 / MXFP4. The smallest real configuration today is 8× RTX PRO 6000 at UpCloud, $7.78 per hour, about $5,676 a month at 730 hours.
Parameters1T42B active per token
Context1M tokensmaximum published
Checkpoint573 GBINT4-packed, as published
Licensemitcheck before commercial use
ReleasedSep 22, 2026documented date
Smallest configuration · INT4 / MXFP4
8× RTX PRO 6000768 GB VRAM · 90% used
- Per hour
- $7.78
- Per month
- $5,676
- Cheapest at
UpCloud
Memory · 1 precision · 6 context lengthsHow much VRAM it needs
Weights plus runtime overhead, then the KV cache that grows with the context you serve.
| Precision | Weights | Weights + headroom | 4k context | 32k context | 128k context | 256k context | 1M context |
|---|---|---|---|---|---|---|---|
| INT4-packedpublished checkpoint | 573 GB | 689 GB | 690 GB | 703 GB | 745 GB | 801 GB | 1,140 GB |
| KV cacheone sequence, FP16 cache, included above | — | — | 2 GB | 15 GB | 57 GB | 113 GB | 451 GB |
Cross-checked on Hugging Face Oct 28, 2026: 1024.2B parameters in the safetensors headers, 573.5 GB of weight files, attention layout from config.json (70 layers, 8 KV heads × 192). Method: weights as published at the native precision, verified on Hugging Face 2026-10-28; other precisions estimated at 1 (8-bit) and 0.6 (4-bit) bytes per parameter; × 1.2 runtime headroom; KV cache = 6144 bytes per token per layer × 70 full-attention layers (FP16 cache, one sequence).config.json ↗
Smallest configuration per GPU · 109 GPUs · cheapest eligible on-demand offer per GPU count, every provider on fileWhat it costs to run, GPU by GPU
The fewest GPUs of one type whose total memory fits the model at short context, priced from the cheapest eligible on-demand offer for exactly that count (45-day window, newest reading per provider), with the next providers beside it. Monthly = 730 hours.
| GPU | Smallest config | VRAM used | $ / hour | $ / month | Cheapest at |
|---|---|---|---|---|---|
| RTX PRO 6000Best fit | 8× 96 GB768 GB total | 90% | $7.78 | $5,676 | UpCloudTrust 75 · read Nov 2 · receipt · also Vast.ai $9.58, Lium $12.80 |
| Gaudi 2 | 8× 96 GB768 GB total | 90% | $8.27 | $6,036 |  Cyfuture AInot scored · read Nov 2 · receipt |
| MI300X | 4× 192 GB768 GB total | 90% | $11.31 | $8,257 | Cyfuture AInot scored · read Nov 2 · receipt · also Hot Aisle $11.96, AtmosCompute $13.38 |
| MI325X | 4× 256 GB1024 GB total | 67% | $12.29 | $8,972 | Cyfuture AInot scored · read Nov 2 · receipt |
| B200 | 4× 180 GB720 GB total | 96% tight | $16.46 | $12,013 | UpCloudTrust 75 · read Nov 2 · receipt · also Deep Infra $19.56, Lambda $27.16 |
| H100 NVL | 8× 94 GB752 GB total | 92% tight | $16.48 | $12,030 |  GPU.ainot scored · read Nov 2 · receipt · also Spheron $21.97, AtmosCompute $35.73 |
| H200 | 8× 141 GB1128 GB total | 61% | $21.04 | $15,359 |  Airon AInot scored · read Nov 2 · receipt · also Geodd $21.76, Lium $23.92 |
| B300 | 4× 270 GB1080 GB total | 64% | $23.29 | $17,004 | UpCloudTrust 75 · read Nov 2 · receipt · also FPT AI Factory $27.96, Verda $45.40 |
| MI355X | 8× 288 GB2304 GB total | 30% | $23.60 | $17,228 |  TensorWaveTrust 64 · read Nov 2 · receipt |
| H200 NVL | 8× 141 GB1128 GB total | 61% | $26.09 | $19,042 |  IONOSnot scored · read Nov 2 · receipt · also Geodd $27.12, Vast.ai $39.47 |
| B200 CC | 4× 180 GB720 GB total | 96% tight | $36.19 | $26,419 |  VerdaTrust 81 · read Nov 2 · receipt |
| B300 CC | 4× 270 GB1080 GB total | 64% | $46.31 | $33,806 | VerdaTrust 81 · read Nov 2 · receipt |
| GB300 | 4× 279 GB1116 GB total | 62% | $52.21 | $38,113 | VerdaTrust 81 · read Nov 2 · receipt |
Throughput estimates · Oct 4, 2026Tokens per second, and what a million costs
Chat and reasoning. Speed estimated from memory bandwidth and model size, never measured. The cost per 1M output tokens is an estimate at the prices read on Oct 4, 2026; it is not recomputed here.
| GPU | Cards | Precision | tok/s per user | per 1M output tokens · estimate |
|---|---|---|---|---|
| MI300X | 4 | FP4 | 36.5 | $2.57 |
| B200 | 4 | FP4 | 63.6 | $3.01 |
| H200 | 8 | FP4 | 40.0 | $4.81 |
| RTX PRO 6000 | 8 | FP4 | 24.5 | $5.78 |
| A100 80GB | 8 | FP4 | 31.6 | $5.96 |
| H100 SXM | 8 | FP4 | 38.9 | $7.57 |
| H100 PCIe | 8 | FP4 | 23.0 | $11.80 |
cards, precision and speed estimated from memory bandwidth and model size; cost recomputed by the site from verified prices · cheapest per output at the prices of Oct 4, 2026: MI300XWhy estimates, and their limits →
MiMo V2.6 Pro RL · related