// AI Platform Drift

DeepSeek Flash Latest completion pricing dropped 34% on OpenRouter

OpenRouter model catalogue · published 2026-09-30 22:00 UTC
WORTH KNOWING BUDGET PRICINGMODEL
WHAT CHANGEDOpenRouter reduced pricing for deepseek/deepseek-flash-latest from $0.60/M to $0.396/M completion tokens (34% decrease) and $0.02/M to $0.0198/M prompt tokens (1% decrease), effective now.
WHY IT MATTERSTotal cost per request drops 33-34% depending on output length. A workload generating ten completion tokens per prompt token now costs 34% less to run.
WHAT TO DOReforecast your OpenRouter spend if you route traffic to deepseek-flash-latest, and notify whoever owns the AI budget that the run rate just dropped by roughly a third.
Model-reported confidence 95% (not measured accuracy) Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-flash-latest
 name: DeepSeek: DeepSeek Flash Latest
-pricing.prompt: 0.00000002
-pricing.completion: 0.0000006
+pricing.prompt: 0.0000000198
+pricing.completion: 0.000000396
 context_length: 1048576
 top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
Triage signals below are model-reported assertions, preserved verbatim; they are not independent arithmetic verification.
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningCompletion token pricing dropped significantly; affects cost calculations for production workloads.
signalspricing.prompt decreased ~1% · pricing.completion decreased ~34% · model pricing change on public catalogue
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)