Teamday

Coding Model · DeepSeek

DeepSeek V4 Pro

DeepSeek V4 Pro is a million-token reasoning and coding model; the GA build 0813 shipped August 13, 2026 with a 1M context, 384K max output, and reasoning effort low/high/max. DeepSeek switched to peak/off-peak billing on August 16, 2026, with off-peak at half price. Teamday runs it through the Pi harness — directly against the DeepSeek API (the best path with a DeepSeek key), or via OpenRouter or Together.

Use DeepSeek V4 Pro in Teamday →All models

At a glance (verified 2026-09-02)

Vendor
DeepSeek
Type
Coding Model
Context window
1M tokens
Pricing
DeepSeek peak (01:00–04:00 and 06:00–10:00 UTC, Mon–Fri): $1.32 input / $0.044 cache-hit / $3.96 output per 1M tokens; off-peak: $0.66 / $0.022 / $1.98.
OpenRouter ID
deepseek/deepseek-v4-pro-0813

Best for

Million-token reasoningLow-cost frontier codingAgentic work on a budget

How to use DeepSeek V4 Pro in Teamday

Pick DeepSeek V4 Pro on Teamday's Pi harness — direct with a DEEPSEEK_API_KEY (recommended), or via OpenRouter or Together.

Harness: pi →Max →James →

Frequently asked questions

How much does DeepSeek V4 Pro cost?

Since August 16, 2026 DeepSeek bills by time of day. Peak (01:00–04:00 and 06:00–10:00 UTC, Monday–Friday): $1.32 per 1M input tokens, $0.044 for cache-hit input, $3.96 per 1M output. Off-peak is half price: $0.66 / $0.022 / $1.98. Together hosts deepseek-ai/DeepSeek-V4-Pro-0813 at $1.32 / $0.13 cached / $3.96; DeepSeek's own OpenRouter endpoint for deepseek/deepseek-v4-pro-0813 applies the same peak and off-peak rates (October 5, 2026); third-party OpenRouter providers price it differently.

What is DeepSeek V4 Pro's context window?

The GA build 0813 has a 1M-token context window and up to 384K output tokens, with reasoning effort selectable at low, high, or max.

Is there a cheaper DeepSeek option?

Yes — DeepSeek V4 Flash (build 0731) costs $0.44 input / $0.014 cache-hit / $1.32 output per 1M tokens at peak and $0.22 / $0.007 / $0.66 off-peak on the direct API, for faster coding, research, and high-volume tasks.

How do I run DeepSeek V4 Pro in Teamday?

Select it on the Pi harness in the chat model picker; the direct DeepSeek API route is the best path when your org has a DeepSeek API key.

Related

DeepSeek V4 Pro is one of the models Teamday's AI employees already run — no API keys, no infrastructure.

Hire your first AI employee →