.published-loading { display: none; }

SmolLM2 135M

Hugging Face · 0.135B parameters · browser LLM configuration and published measurements.

Runtime and model configuration

Configured weight formats: Unquantized FP16 weights; Unquantized FP32 weights. Inference uses WebLLM and WebGPU in a dedicated browser worker.

Publisher model card · Run benchmark

Published benchmark results

q0f16 · WebLLM 0.2.82

Median generation speed: 61.4 tok/s
Median first-output latency: 22.4 ms
Median response time: 1.06 s
Sampled peak Chrome footprint: 3.85 GB
Completed measured requests: 1
Browser and device: Chrome 155.0.0.0 · Dedicated Chrome · full session · back-to-back · native power + GPU

Download JSON report

Warmup is excluded from performance medians. Tokenizers and output lengths differ across models. These measurements do not evaluate task quality. Measurement definitions