InferenceX
中文

Inference Cost per Million Tokens

Hyperscaler cost ↓ Lower is better Source: InferenceX & SemiAnalysis Market July 2026 AI Cloud TCO Model

  • DeepSeek V4 Pro 1.6T

    8K/1K

    B200
    $0.16266% cheaper than this platform’s Jul 12 result
    Dynamo vLLM · FP4
    Compare curves
    MI355X
    $0.53495% cheaper than this platform’s Jul 10 result
    SGLang · FP4
    Compare curves
    B300
    $0.18046% cheaper than this platform’s Jul 12 result
    Dynamo SGLang · FP4
    Compare curves
    GB200 NVL72
    $0.864
    Dynamo vLLM · FP4
    GB300 NVL72
    $0.77137% cheaper than this platform’s Jul 6 result
    Dynamo SGLang · FP4STP
    Compare curves

  • DeepSeek V4 Pro 1.6T

    Long Context Multi-Turn Realistic Agentic Scenario (AgentX)

    B200
    $0.077
    vLLM · FP4
    MI355X
    cannot reach @75
    B300
    $0.047
    vLLM · FP4
    GB200 NVL72
    no data for this scenario
    GB300 NVL72
    $0.017
    Dynamo vLLM · FP4

  • Kimi K3 2.8T

    Long Context Multi-Turn Realistic Agentic Scenario (AgentX)

    B200
    cannot reach @75
    MI355X
    no data for this scenario
    B300
    $0.177
    vLLM · FP4
    GB200 NVL72
    $0.597
    Dynamo vLLM · FP4
    GB300 NVL72
    no data for this scenario

  • MiniMax M3 428B

    8K/1K

    B200
    $0.056About the same cost as this platform’s Jul 6 result
    vLLM · FP4
    Compare curves
    MI355X
    $0.123
    vLLM · FP4
    B300
    $0.074About the same cost as this platform’s Jul 6 result
    vLLM · FP4
    Compare curves
    GB200 NVL72
    $0.18515% cheaper than this platform’s Jun 29 result
    Dynamo vLLM · FP8STP
    Compare curves
    GB300 NVL72
    $0.2126% cheaper than this platform’s Jun 26 result
    Dynamo vLLM · FP8STP
    Compare curves

  • MiniMax M3 428B

    Long Context Multi-Turn Realistic Agentic Scenario (AgentX)

    B200
    $0.012
    vLLM · FP4
    MI355X
    no data for this scenario
    B300
    $0.012
    vLLM · FP4
    GB200 NVL72
    no data for this scenario
    GB300 NVL72
    no data for this scenario

  • GLM5.2

    Long Context Multi-Turn Realistic Agentic Scenario (AgentX)

    B200
    $0.073
    SGLang · FP4
    MI355X
    no data for this scenario
    B300
    no exact @75 result
    GB200 NVL72
    no data for this scenario
    GB300 NVL72
    no data for this scenario

  • Qwen3.5 397B

    8K/1K

    B200
    $0.053About the same cost as this platform’s Jul 5 result
    SGLang · FP4
    Compare curves
    MI355X
    $0.08230% cheaper than this platform’s Jun 15 result
    SGLang · FP4
    Compare curves
    B300
    $0.0507% cheaper than this platform’s Jul 4 result
    SGLang · FP4
    Compare curves
    GB200 NVL72
    $0.314
    Dynamo SGLang · FP8STP
    GB300 NVL72
    $0.153About the same cost as this platform’s Jun 25 result
    Dynamo SGLang · FP4STP
    Compare curves

  • Qwen3.5 397B

    Long Context Multi-Turn Realistic Agentic Scenario (AgentX)

    B200
    $0.013
    SGLang · FP8
    MI355X
    no data for this scenario
    B300
    $0.008
    SGLang · FP4
    GB200 NVL72
    no data for this scenario
    GB300 NVL72
    $0.009
    Dynamo SGLang · FP4

Current cost and change versus the latest validated platform result 30–60 days earlier.

Platforms without a valid 30-day comparison show current cost only.

If a chip does not have FP4 spec decoding available, the next best available configuration is used.