DeepSeek V4.1 Flash (fast)

PRICE / USD

$0.3

Per 1 million output tokens · input billed separately

Input / 1M tokens

$0.3

Output / 1M tokens

$1.2

Cached input / 1M tokens

$0.006

Cache write / 1M tokens

Not verified

Off-peak rates are 50% lower. Peak: weekdays 01:00–04:00 and 06:00–10:00 UTC.

SPEED & CAPABILITIES

214.1 tokens/s

DeepSeek API · V4.1 Flash · Max reasoning. Artificial Analysis reports a rolling 72-hour median of output speed after the first token, not total response time. Checked 2026-10-05; speed varies with workload and serving conditions.

Artificial Analysis · performance methodology v2.2.0 ↗ · Checked 2026-10-05. (fast) means median output above 150 tokens/s in these settings.

Model score

39

Our own output-speed measurement

Not measured

Context window

1,000,000

API model ID

deepseek-flash

Artificial Analysis tested DeepSeek V4.1 Flash · max. This mostly English, text-based index is not a guarantee of performance on your workload. Artificial Analysis · DeepSeek V4.1 Flash · max ↗

The old V4 Flash API aliases now serve V4.1 Flash. Preview benchmark scores are not transferred to this version.

Suggested starting role: everyday coding. A low-cost starting point for code changes and tool workflows, based on its documented reasoning and agent support. Check your own tests before using it for larger changes. This is editorial guidance, not a benchmark score.