DeepSeek V4 Pro 1.6T
8K/1K
Hyperscaler cost ↓ Lower is better Source: InferenceX & SemiAnalysis Market July 2026 AI Cloud TCO Model
| Model · Scenario | B200 · Reference | MI355X | B300 | GB200 NVL72 | GB300 NVL72 |
|---|---|---|---|---|---|
DeepSeek V4 Pro 1.6T8K/1K | vLLM · FP4 | $0.13638% cheaper than B200 SGLang · FP4 | $0.18218% cheaper than B200 vLLM · FP4 | $0.225About the same cost as B200 Dynamo vLLM · FP4 | — no exact @30 result |
DeepSeek V4 Pro 1.6TLong Context Multi-Turn Realistic Agentic Scenario (AgentX) | vLLM · FP4 | $0.071218% more expensive than B200 vLLM · FP4 | $0.0248% more expensive than B200 vLLM · FP4 | — no data for this scenario | — no exact @30 result |
Kimi K3 2.8TLong Context Multi-Turn Realistic Agentic Scenario (AgentX) | Dynamo vLLM · FP4 · STP | — no data for this scenario | $0.13587% cheaper than B200 vLLM · FP4 | $0.42959% cheaper than B200 Dynamo vLLM · FP4 | — no data for this scenario |
MiniMax M3 428B8K/1K | vLLM · FP8 | $0.10888% more expensive than B200 vLLM · FP4 | $0.04620% cheaper than B200 vLLM · FP4 | $0.08649% more expensive than B200 Dynamo vLLM · FP8 · STP | — no exact @30 result |
MiniMax M3 428BLong Context Multi-Turn Realistic Agentic Scenario (AgentX) | — no exact @30 result | — no data for this scenario | — no exact @30 result | — no data for this scenario | — no data for this scenario |
GLM5.2Long Context Multi-Turn Realistic Agentic Scenario (AgentX) | — no exact @30 result | — no data for this scenario | — no exact @30 result | — no data for this scenario | — no data for this scenario |
Qwen3.5 397B8K/1K | SGLang · FP8 | $0.057About the same cost as B200 SGLang · FP4 | $0.06923% more expensive than B200 SGLang · FP8 | — no exact @30 result | — no exact @30 result |
Qwen3.5 397BLong Context Multi-Turn Realistic Agentic Scenario (AgentX) | — no exact @30 result | — no data for this scenario | — no exact @30 result | — no data for this scenario | $0.019No B200 baseline to compare against Dynamo SGLang · FP4 |
DeepSeek R1 0528 671BMaintenance8K/1K | SGLang · FP4 | $0.367420% more expensive than B200 MoRI SGLang · FP8 | $0.284302% more expensive than B200 SGLang · FP8 | — no exact @30 result | — no exact @30 result |
Kimi K2.5/2.6/2.7-Code 1TDeprecated8K/1K | vLLM · FP4 · STP | $0.105About the same cost as B200 vLLM · FP4 · STP | $0.12422% more expensive than B200 vLLM · FP4 · STP | — no exact @30 result | — no exact @30 result |
GLM5/5.1 744BDeprecated8K/1K | SGLang · FP4 | $0.283115% more expensive than B200 SGLang · FP8 | $0.16324% more expensive than B200 SGLang · FP4 | — no exact @30 result | — no exact @30 result |
gpt-oss 120BDeprecated8K/1K | — no exact @30 result | — no exact @30 result | — no data for this scenario | — no data for this scenario | — no data for this scenario |
MiniMax M2.5/2.7 230BDeprecated8K/1K | vLLM · FP4 · STP | $0.073150% more expensive than B200 vLLM · FP8 · STP | $0.03725% more expensive than B200 vLLM · FP4 · STP | — no exact @30 result | $0.04245% more expensive than B200 Dynamo vLLM · FP4 · STP |
Llama 3.3 70B InstructDeprecated8K/1K | vLLM · FP4 · STP | $0.07845% more expensive than B200 vLLM · FP4 · STP | — no data for this scenario | — no data for this scenario | — no data for this scenario |
8K/1K
Long Context Multi-Turn Realistic Agentic Scenario (AgentX)
Long Context Multi-Turn Realistic Agentic Scenario (AgentX)
8K/1K
Long Context Multi-Turn Realistic Agentic Scenario (AgentX)
Long Context Multi-Turn Realistic Agentic Scenario (AgentX)
8K/1K
Long Context Multi-Turn Realistic Agentic Scenario (AgentX)
8K/1K
8K/1K
8K/1K
8K/1K
8K/1K
8K/1K
— = no result. ∞ = B200 baseline unavailable.
If a chip does not have FP4 spec decoding available, the next best available configuration is used.