DeepSeek V4.1 Flash 1017 Doubles Its API Price: $0.15→$0.30 Input, $0.60→$1.20 Output

The Change

DeepSeek V4.1 Flash 1017 has repriced at a clean +100% across the board. Input tokens move from $0.15 to $0.30 per 1M tokens; output tokens jump from $0.60 to $1.20 per 1M tokens. No partial adjustment — everything doubled.

See the full pricing history on the DeepSeek V4.1 Flash 1017 — live specs & price history page.

Does It Matter for Your Workload?

At the old rate, a workload burning 10M output tokens a month cost $6.00. At $1.20 per 1M, that same workload now runs $12.00. High-volume, output-heavy pipelines — summarisation, code generation, agentic loops — feel this most. Light or infrequent API users will barely notice.

The model still sits in the budget tier relative to frontier models, but a 100% increase in one move is unusual and worth factoring into any cost projection you built before today.

What to Do Next

• Recalculate your monthly token spend using the new figures: $0.30 input / $1.20 output per 1M tokens.
• Compare alternatives — our model comparison tool lets you filter by input/output price to find models at the old DeepSeek Flash price point or lower.
• Watch for further movement. A single doubling can signal ongoing repricing; it's worth setting a price-change alert if you're locked into this model for production use.

Bottom line: if DeepSeek V4.1 Flash 1017 was in your stack because of its price-to-performance ratio, that ratio just got meaningfully worse. Worth a five-minute audit before your next billing cycle.

Ezra, Scout AI Team