Kimi K3.1 Pro: Self-Hosting Estimate
Kimi K3.1 Pro (Moonshot AI) on different GPUs: configuration, estimated performance and use cases.
Data updated: 2026-09-16B300 288GB
288 GB · 8,000 GB/s · Top-tier Blackwell card (no official channels to China)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 8priced as 1 × 8-GPU HGX node
- Est. decode speed
- 1,231 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 2,092 tok/sin 11,250 tok/s · many concurrent requests
- Hardware list price
- $299,058–$498,429Cloud rental (live) ×8: $63.08–$63.08/hr (2026-11-04)
- Hardware market price
- $1,047,167
- Typical use
- Enterprise-grade ultra-large-scale training and high-concurrency inference clusters
GB200 192GB (single card equivalent)
192 GB · 8,000 GB/s · Top-tier Blackwell card (no official channels to China)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 16 (exceeds one node: multi-node needed)
- Est. decode speed
- 2,462 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 4,185 tok/sin 25,000 tok/s · many concurrent requests
- Hardware list price
- $479,088–$718,632
- Hardware market price
- $1,436,115–$2,872,231
- Typical use
- Ultra-large-scale trillion-parameter model training and inference clusters (networked with Grace CPU)
B200 180GB
180 GB · 8,000 GB/s · Top-tier Blackwell card (no official channels to China)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 16 (exceeds one node: multi-node needed)priced as 2 × 8-GPU HGX node
- Est. decode speed
- 2,462 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 4,185 tok/sin 22,500 tok/s · many concurrent requests
- Hardware list price
- $598,116–$996,858Cloud rental (live) ×16: $119.17–$200.16/hr (2026-11-05)
- Hardware market price
- $897,572–$1,047,167
- Typical use
- Enterprise-grade large-scale online inference and training
H200 141GB
141 GB · 4,800 GB/s · Top-tier export-controlled card (channels to China are volatile)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 16 (exceeds one node: multi-node needed)priced as 2 × 8-GPU HGX node
- Est. decode speed
- 1,477 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 2,511 tok/sin 9,890 tok/s · many concurrent requests
- Hardware list price
- $559,673–$999,417Cloud rental (live) ×16: $47.86–$73.71/hr (2026-11-05)
- Hardware market price
- $448,786–$466,737
- Typical use
- Enterprise-grade production environment high-concurrency inference
H100 SXM5 80GB
80 GB · 3,350 GB/s · Top-tier export-controlled card (no official channels to China)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)priced as 4 × 8-GPU HGX node
- Est. decode speed
- 2,062 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 3,505 tok/sin 19,780 tok/s · many concurrent requests
- Hardware list price
- $959,440–$3,326,057Cloud rental (live) ×32: $61.89–$131.62/hr (2026-11-05)
- Hardware market price
- $1,282,930
- Typical use
- Enterprise-grade production environment high-concurrency inference/training
H800 SXM 80GB
80 GB · 3,350 GB/s · Top-tier export-controlled card (discontinued)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)priced as 4 × 8-GPU HGX node
- Est. decode speed
- 2,062 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 3,505 tok/sin 19,780 tok/s · many concurrent requests
- Hardware list price
- $1,172,828–$1,244,633
- Hardware market price
- $1,615,630–$1,675,468
- Typical use
- Enterprise-grade production deployment (original China-compliant model, now discontinued)
H20 96GB
96 GB · 4,000 GB/s · Export-compliant special supply card (significantly reduced compute power)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)
- Est. decode speed
- 2,462 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 4,185 tok/sin 2,960 tok/s · many concurrent requests
- Hardware list price
- n/a
- Hardware market price
- $382,964–$526,576
- Typical use
- Suboptimal choice in export-compliant environments, medium-to-large scale inference deployment (FP16 compute power is only about 15% of H100, but higher memory bandwidth offers an advantage in decode-intensive scenarios)
A800 PCIe 80GB
80 GB · 1,935 GB/s · Compliant special supply card (export compliance status adjusted multiple times)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,191 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,215 tok/sin 3,744 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $478,705–$497,853
- Hardware market price
- $517,002–$574,446
- Typical use
- Medium-scale inference/training deployment in export-compliant environments
Ascend 910C 128GB
128 GB · 3,200 GB/s · Huawei dual-chip data center card (domestically produced)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 16 (exceeds one node: multi-node needed)
- Est. decode speed
- 985 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,674 tok/sin 7,520 tok/s · many concurrent requests
- Hardware list price
- n/a
- Hardware market price
- $191,482–$478,705
- Typical use
- Domestically produced large model training and inference clusters, CloudMatrix and other super-node solutions
Ascend 910B 64GB
64 GB · 400 GB/s · Domestic alternative (not subject to US export controls)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)
- Est. decode speed
- 246 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 418 tok/sin 6,400 tok/s · many concurrent requests
- Hardware list price
- n/a
- Hardware market price
- $526,576–$574,446
- Typical use
- Domestic alternative, not subject to US export controls, medium-to-large scale inference/training deployment, memory bandwidth significantly lower than NVIDIA cards of similar VRAM capacity
RTX PRO 6000 Blackwell 96GB
96 GB · 1,792 GB/s · Professional workstation/server card (China-specific version is 6000D 84GB)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,103 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,125 tok/sin 6,000 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $512,798Cloud rental (live) ×32: $38.50–$72.61/hr (2026-11-05)
- Hardware market price
- $562,766
- Typical use
- Single-card large VRAM inference, local deployment of medium MoE models, private all-in-one machine
RTX PRO 5500 Blackwell 84GB
84 GB · 1,398 GB/s · Professional workstation card (specs are NVIDIA preliminary data)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 32 (exceeds one node: multi-node needed)
- Est. decode speed
- 860 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 878 tok/sin 5,424 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $364,587
- Hardware market price
- $406,899–$430,835
- Typical use
- Workstation large VRAM inference, departmental private deployment
RTX PRO 5000 Blackwell 48GB
48 GB · 1,344 GB/s · Professional workstation card- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 64 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,654 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,687 tok/sin 6,192 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- n/aCloud rental (live) ×64: $35.90–$68.42/hr (2026-11-05)
- Hardware market price
- $517,002
- Typical use
- Small to medium model inference, development testing, single-machine multi-card private deployment
L40S 48GB
48 GB · 864 GB/s · Compliant inference card- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 64 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,063 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,085 tok/sin 8,688 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $430,835–$622,317Cloud rental (live) ×64: $38.46–$58.37/hr (2026-11-05)
- Hardware market price
- $430,835–$622,317
- Typical use
- SME teams building inference services, medium-low concurrency
L20 48GB
48 GB · 864 GB/s · Compliant inference card- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 64 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,063 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,085 tok/sin 2,868 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $223,527–$224,809
- Hardware market price
- $335,094
- Typical use
- SME teams building inference services, medium-low concurrency
RTX 5090 32GB
32 GB · 1,792 GB/s · Consumer flagship (not sold in China, 5090D for domestic market)- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 64 (exceeds one node: multi-node needed)
- Est. decode speed
- 2,206 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 2,250 tok/sin 10,056 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- n/aCloud rental (live) ×64: $28.29–$34.05/hr (2026-11-05)
- Hardware market price
- $430,825
- Typical use
- Local inference and development for individuals/small teams, multi-card VRAM pooling for quantized MoE
RTX 5090D 24GB
24 GB · 1,344 GB/s · Consumer-grade/small-scale self-built- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 70 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,809 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,845 tok/sin 7,796 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $172,772Cloud rental (live) ×70: $30.94–$37.10/hr (2026-11-05)
- Hardware market price
- $293,207–$418,867
- Typical use
- Personal development/small-scale verification/POC, not recommended for production-level high concurrency
RTX 4090D 24GB
24 GB · 1,008 GB/s · Consumer-grade/small-scale self-built- Suggested precision
- INT4 (AWQ/GPTQ)
- GPUs needed
- 70 (exceeds one node: multi-node needed)
- Est. decode speed
- 1,357 tok/ssingle-stream theoretical ceiling
- Est. batched throughput
- out 1,384 tok/sin 7,718 tok/s · many concurrent requests · PCIe multi-GPU penalty
- Hardware list price
- $136,121–$198,951Cloud rental (live) ×70: $23.66–$30.52/hr (2026-11-05)
- Hardware market price
- $272,264–$314,150
- Typical use
- Personal development/small-scale verification/POC, not recommended for production-level high concurrency
FAQ
FAQ
How is the estimated decode speed calculated?
Estimated speed = (memory bandwidth × GPU count) ÷ (active parameters × bytes per parameter): the memory-bound upper limit for one request (roofline). It scales linearly with GPU count and ignores interconnect overhead, batching and framework losses, so real results are usually well below it.
Can I trust the speed numbers for consumer GPUs?
The linear assumption is reasonable for NVLink data-center cards (H/B series) but too optimistic for consumer PCIe cards without NVLink (RTX 5090D/4090D); multi-GPU consumer setups lose far more to communication, so equal-looking numbers do not mean equal performance.
How is “Suggested precision” chosen?
It is the highest precision that fits within 8 GPUs; beyond that it is flagged as multi-node.
Where do the prices come from?
USD converted at ¥6.68 per USD. Cloud rental is the international vast.ai market rate, not a China purchase price.