1.2x faster median latency on H100, vLLM 0.25.1, concurrency 8, recorded Jul 27, 2026.
1 measured package. Results apply to the recorded conditions.
Packages are no longer sold. For hosted inference, see Model APIs.
| Package | GPU | Engine | Optimization | Verified |
|---|---|---|---|---|
| H100, Channelwise FP8 weights, dynamic per-token activationsarex-turbo-fp8cd-h100-vllm | GPUH100 | EnginevLLM 0.25.1 | OptimizationChannelwise FP8 weights, dynamic per-token activations | VerifiedJul 27, 2026 |
Channelwise FP8 changes the weights without reading calibration data, while activation scales are chosen per token at runtime.
Published Jul 27, 2026. Proof verified Jul 27, 2026.
chat-128
vLLM 0.25.1. Concurrency 8. Proof verified Jul 27, 2026.
Median latency measured 666 ms at baseline and 551 ms optimized.
gsm8k no measurable accuracy change, passed.
Published package measurements for AREX-Turbo, grouped from the public RunInfra catalog.
Measured by RunInfra on H100 with vLLM 0.25.1; proof dates are listed with each package.
We report package measurements here, and each package remains subject to its listed license.
Citation: RunInfra (2026). Published package measurements and signed kit records. https://runinfra.ai/model.
Other GPU measurements for this model not published.
Other engine measurements for this model not published.
Measurements for context lengths not listed in these packages not published.
No. Package sales are closed. For hosted inference, see Model APIs at runinfra.ai/inference-api.
H100, Channelwise FP8 weights, dynamic per-token activations is measured on H100 with vLLM 0.25.1. Proof verified Jul 27, 2026.
gsm8k no measurable accuracy change, passed. Proof verified Jul 27, 2026.
Use a workspace API key and pay for input, cached input, and output tokens.
View Model APIs