DeepSeek-V4 API Pricing Leaks: 1M Context, Peak/Off-Peak Billing
aigclink · x · 2026-08-13
Leaked DeepSeek API documentation reveals detailed specs and pricing for the upcoming DeepSeek-V4 series (Flash and Pro).
- Core Specs: Supports a massive 1M context length with up to 384K output. Fully supports reasoning modes, Tool Calls, Responses API, and Anthropic API compatibility.
- Current Pricing: Flash model costs 1 RMB/Mtok input (cache miss) and 2 RMB/Mtok output; Pro model is 3 RMB input and 6 RMB output.
- Peak/Off-Peak Billing: Starting August 2026, DeepSeek will introduce peak/off-peak pricing, with off-peak rates at 50% of peak rates (e.g., Flash peak output at 9 RMB, off-peak at 4.5 RMB).
Related event: DeepSeek Launches V4-Pro Model with Massive API Price Hikes(47 posts)→
More from Models
- DeepSeek Drops Open-Weight V4-Pro Model with Million-Token Context — multimodalart · 2026-08-13
- User Complains Ollama Pro Subscription Feels Like a Scam: $200 Paid, Limits Lowered, No Refund — cheapybastard · 2026-08-13
- User Reports Grok 4.6 as Highly Pedantic with Extensive Self-Checking — intellectronica · 2026-08-13
- Krea, FLUX and Others Promise Open Weights, But How Meaningful Are They? — felixsanz · 2026-08-13
- Guardrail Differences: OpenAI Hacks, While Gemini Declares You Socially Dead — MakeSureRegs · 2026-08-13
- Gemini 3.7 Flash spotted on Google Cloud Console, release imminent — Rare_Bunch4348 · 2026-08-13