Qwen3.6 27B API Prices Rise 108% — Input Cost Jumps to $0.60/M Tokens

What Changed

Qwen has increased API pricing on Qwen3.6 27B across both token directions:

• Input: $0.289 → $0.60 per 1M tokens (+108%)
• Output: $2.40 → $3.60 per 1M tokens

The input price increase is the more aggressive move — a clean doubling-plus that will hit read-heavy workloads like RAG pipelines, document classification, and long-context summarization hardest. Output pricing rose a more moderate 50%, so generation-heavy tasks (drafting, code completion) feel less of the squeeze in relative terms, though $3.60/M output is no longer a budget tier.

Does This Matter for Your Workload?

It depends on your token ratio. A workload processing 10M input tokens monthly now costs roughly $3,110 more per year at the new rate — not trivial for indie developers or startups running at scale. Enterprises with negotiated contracts may be insulated short-term, but anyone on pay-as-you-go should audit usage now.

If most of your spend sits on output tokens, the impact is proportionally smaller. But if you're feeding large contexts in and getting short answers back, this increase lands hard.

What to Do

Check the current rate card and full spec history on the Qwen3.6 27B — live specs & price history page before adjusting budgets. It's also worth benchmarking comparable mid-size open-weight models on our site — several remain priced well below $0.60/M input and may cover similar use cases depending on your quality requirements.

Price increases on capable mid-size models aren't unusual as providers move past introductory rates, but a +108% input jump is steep enough to warrant a real alternatives review rather than a passive rollover.

Ezra, Scout AI Team