DGX Spark benchmarks
Model catalog
Filter by hardware and serving engine. Configurations are grouped by model so you can compare runs directly.
30 configurations across 14 models
Decode sweep · ~256 input / 512 output tokens per request
No configurations match these filters.
DeepSeek V4.1 Flash 1017 EXL3-EAQ
2 configurations
2× DGX Spark
TensorFold
- Single-stream decode
- 67.50output tok/s · c1
- Sustained throughput
- 108.73output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval incompleteSafety gate passed
2× DGX Spark
TensorFold
- Single-stream decode
- 65.62output tok/s · c1
- Sustained throughput
- 103.17output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval incompleteSafety gate passed
GLM-5.3 Flash
7 configurations
2× DGX Spark
TensorFold
- Single-stream decode
- 41.85output tok/s · c1
- Sustained throughput
- 65.01output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate failed
2× DGX Spark
TensorFold
- Single-stream decode
- 42.67output tok/s · c1
- Sustained throughput
- 62.29output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate failed
2× DGX Spark
TensorFold
- Single-stream decode
- 54.53output tok/s · c1
- Sustained throughput
- 80.18output tok/s · c8 · 20 min
- Longest tested prompt
- 131,072target input tokens · c1
Configured 1,048,576 tokens ·Validated Unavailable
Tool eval completeSafety gate failed
2× DGX Spark
vLLM
- Single-stream decode
- 14.20output tok/s · c1
- Sustained throughput
- 60.90output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
TensorFold
- Single-stream decode
- 51.49output tok/s · c1
- Sustained throughput
- 84.90output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
vLLM
- Single-stream decode
- 13.96output tok/s · c1
- Sustained throughput
- 21.88output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 850,000 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
vLLM
- Single-stream decode
- 35.13output tok/s · c1
- Sustained throughput
- 38.53output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,000,000 tokens ·Validated ≈240K passkey
Tool eval unavailable
Qwen3.8 27B
3 configurations
1× DGX Spark
TensorFold
- Single-stream decode
- 50.25output tok/s · c1
- Sustained throughput
- 160.59output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval incompleteSafety gate failed
1× DGX Spark
SGLang
- Single-stream decode
- 31.83output tok/s · c1
- Sustained throughput
- 157.80output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval unavailable
1× DGX Spark
SGLang
- Single-stream decode
- 18.60output tok/s · c1
- Sustained throughput
- 63.89output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval unavailable
Qwen3.8-Flash-Next NVFP4
8 configurations
1× DGX Spark
TensorFold
- Single-stream decode
- 41.54output tok/s · c1
- Sustained throughput
- 77.79output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate failed
1× DGX Spark
TensorFold
- Single-stream decode
- 34.29output tok/s · c1
- Sustained throughput
- 85.72output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval incompleteSafety gate passed
2× DGX Spark
TensorFold
- Single-stream decode
- 83.45output tok/s · c1
- Sustained throughput
- 432.90output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval incompleteSafety gate passed
2× DGX Spark
vLLM
- Single-stream decode
- 44.62output tok/s · c1
- Sustained throughput
- 196.70output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
vLLM
- Single-stream decode
- 47.34output tok/s · c1
- Sustained throughput
- 197.42output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
vLLM
- Single-stream decode
- 47.51output tok/s · c1
- Sustained throughput
- 201.58output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,000,000 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
vLLM
- Single-stream decode
- 48.93output tok/s · c1
- Sustained throughput
- 197.71output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,000,000 tokens ·Validated ≈240K passkey
Tool eval unavailable
2× DGX Spark
SGLang
- Single-stream decode
- 38.43output tok/s · c1
- Sustained throughput
- 157.61output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval unavailable
DeepSeek V4 Flash 0731
1 configuration
1× DGX Spark
vLLM + SparkInfer
- Single-stream decode
- 29.62output tok/s · c1
- Sustained throughput
- 26.89output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 384,000 tokens ·Validated ≈240K passkey
Tool eval unavailable
DeepSeek V4 Flash DSpark
1 configuration
2× DGX Spark
vLLM (Anemll DSpark)
- Single-stream decode
- 42.12output tok/s · c1
- Sustained throughput
- 111.10output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate passed
DeepSeek V4 Flash Vision-Exp
1 configuration
2× DGX Spark
vLLM (Anemll DSpark 0.1.1)
- Single-stream decode
- 25.93output tok/s · c1
- Sustained throughput
- 84.53output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate failed
DeepSeek V4.1 Flash EXL3
1 configuration
2× DGX Spark
vLLM
- Single-stream decode
- 27.42output tok/s · c1
- Sustained throughput
- 35.31output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 600,000 tokens ·Validated ≈240K passkey
Tool eval unavailable
MiniCPM-V 4.7 35B-A3B
1 configuration
1× DGX Spark
ExLlamaV3
- Single-stream decode
- 47.74output tok/s · c1
- Sustained throughput
- 51.35output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval incompleteSafety gate failed
NousResearch Arete-27B
1 configuration
1× DGX Spark
SGLang
- Single-stream decode
- 4.29output tok/s · c1
- Sustained throughput
- 31.28output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate passed
NVIDIA Nemotron 3.5 Bolt 74B-A9B
1 configuration
1× DGX Spark
TensorFold
- Single-stream decode
- 79.23output tok/s · c1
- Sustained throughput
- 145.52output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate failed
NVIDIA Nemotron 3.5 Super VL 120B-A12B
1 configuration
1× DGX Spark
vLLM
- Single-stream decode
- 34.15output tok/s · c1
- Sustained throughput
- 33.90output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 1,048,576 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate failed
Qwen3.6-35B-A3B NVFP4 (Unsloth)
1 configuration
1× DGX Spark
vLLM
- Single-stream decode
- 77.46output tok/s · c1
- Sustained throughput
- 286.66output tok/s · c8 · 20 min
- Longest tested prompt
- 131,072target input tokens · c1
Configured 262,144 tokens ·Validated Unavailable
Tool eval completeSafety gate failed
Recursive Flint 40B-A2.5B NVFP4 GGUF
1 configuration
1× DGX Spark
llama.cpp
- Single-stream decode
- 81.73output tok/s · c1
- Sustained throughput
- 74.62output tok/s · c8 · 20 min
- Longest tested prompt
- 245,760target input tokens · c1
Configured 262,144 tokens ·Validated ≈240K passkey
Tool eval completeSafety gate passed