Qwen3 VL 30B A3B Instruct API Prices Rise 15% — What It Costs You Now

The Numbers

Qwen has pushed through a +15% price increase on the Qwen3 VL 30B A3B Instruct API. The new rates:

• Input: $0.13 → $0.15 per 1M tokens
• Output: $0.52 → $0.60 per 1M tokens

In absolute terms these are still modest numbers, but the direction matters — this is a straightforward cost increase with no announced capability upgrade attached.

Does It Move the Needle for Your Workload?

For low-volume or experimental use, the delta is negligible — a few extra cents per million tokens won't break a budget. The math gets more meaningful at scale. If you're currently pushing 10M output tokens per month, your monthly output bill climbs from $5.20 to $6.00. At 100M tokens, that's an extra $80/month on output alone.

Heavy vision-language pipelines — document parsing, image captioning at volume, multimodal RAG — are where this 15% bump is most likely to be felt. If output tokens dominate your usage pattern (they usually do in generative tasks), the $0.52 → $0.60 move is the one to watch.

What to Do Next

If this model is a significant line item, now is a reasonable moment to benchmark alternatives. Check the Qwen3 VL 30B A3B Instruct — live specs & price history page for a full timeline of rate changes and to compare against competing vision-language models on our site before your next billing cycle.

If you're not price-sensitive at current volumes, no immediate action needed — but set a usage alert so a volume spike doesn't catch you off guard at the new rates.

Ezra, Scout AI Team