llm-benchmarks
githubdrose.io

Qwen3 VL 30B A3B Instruct Benchmarks


Measured on alibaba

180 runs
Avg Tokens / Second
72.17
Avg Time to First Token (ms)
920.00
Runs Analysed
180
Last Updated
Sep 1, 2026, 04:05 AM

Throughput distribution

Throughput over time

Qwen3 VL 30B A3B Instructalibaba1tok/s

Generated throughput, including reasoning tokens — so a thinking model's line does not collapse; the table ranks on visible tokens · one cell per model, showing its best-covered provider · shared vertical scale, 0 to 1 tok/s (99th percentile) · dashed rule is the model's own mean · 1 samples above the ceiling drawn at it

Every provider serving this model

3 measured
ProviderModelAvg Toks/SecMinMaxAvg TTF (ms)
alibabaQwen3 VL 30B A3B Instruct124.00124.00124.00870.00
alibabaQwen3 VL 30B A3B Instruct47.2042.5050.40870.00
alibabaQwen3 VL 30B A3B Instruct45.3030.10121.001020.00

Frequently Asked Questions

How fast is Qwen3 VL 30B A3B Instruct?
The latest rolling average throughput is 72.17 tokens per second with an average time to first token of 920.00 ms across 180 recent runs.
How often are these benchmarks updated?
Benchmarks refresh automatically whenever the monitoring cron runs. The most recent run completed on Sep 1, 2026, 04:05 AM.

Related models