vLLM hits 130k tok/s on DeepSeek V4 Pro in AgentX benchmark
AccBalanced · x · 2026-08-27
SemiAnalysis released AgentX 1.0, an open-source multi-turn agentic coding benchmark from $3M of real traces. vLLM showed competitive performance: DeepSeek V4 Pro reached 130,093 tok/s/chip, MiniMax M3 77,079 tok/s/chip, and Kimi K3 12,479 tok/s/chip. This highlights vLLM's performance on real-world agentic workloads.
More from Infra
- SGLang ships Day-0 support for FLUX 3 Image Klein: 10 reference images, BF16/FP8/NVFP4 variants — ying11231 · 2026-10-16
- NVIDIA's Nemotron 3.5 Super VL hits CoreWeave: 12B-active multimodal MoE at $0.25/M input tokens — wandb · 2026-10-16
- Chip equipment beats bull case: 2026 WFE seen at $153B, $278B by 2030 — BenBajarin · 2026-10-16
- Steam now launches in 5.23s, beating Epic's 5.47s — and vLLM gets roasted for slow startup — AlpinDale · 2026-10-16
- Shadeform to Announce Six New Features on Oct 19: Lower GPU Costs, More Suppliers — _jaydeepkarale · 2026-10-16
- Fastino launches Training API for SFT/RL, claiming 25-85% cost cut vs Tinker — kalyan_kpl · 2026-10-16