Teamday

Coding Model · DeepSeek

DeepSeek V4 Flash

DeepSeek V4 Flash (build 0731, released July 31, 2026) is the lower-cost DeepSeek V4 tier for faster coding, research, and high-volume tasks, with a 1M context, 384K max output, and reasoning effort low/high/max. DeepSeek bills it at peak/off-peak rates since August 16, 2026, with off-peak at half price. Teamday runs it through the Pi harness against the direct DeepSeek API.

Use DeepSeek V4 Flash 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): $0.44 input / $0.014 cache-hit / $1.32 output per 1M tokens; off-peak: $0.22 / $0.007 / $0.66.

Best for

Faster codingResearchHigh-volume tasksVery low cost

How to use DeepSeek V4 Flash in Teamday

Pick DeepSeek V4 Flash direct on Teamday's Pi harness with a DEEPSEEK_API_KEY.

Harness: pi →James →Sarah →

Frequently asked questions

How much does DeepSeek V4 Flash cost?

On DeepSeek's direct API, peak hours (01:00–04:00 and 06:00–10:00 UTC, Monday–Friday) cost $0.44 per 1M input tokens, $0.014 for cache-hit input, and $1.32 per 1M output; off-peak is half price at $0.22 / $0.007 / $0.66. Peak/off-peak billing started August 16, 2026.

What is DeepSeek V4 Flash best for?

Faster coding, research, and high-volume tasks where V4 Pro's extra capability is not needed.

What is DeepSeek V4 Flash's context window?

Build 0731 has a 1M-token context window and up to 384K output tokens, with reasoning effort selectable at low, high, or max.

How do I use DeepSeek V4 Flash in Teamday?

Select the DeepSeek V4 Flash direct route on the Pi harness with your DeepSeek API key.

Related

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

Hire your first AI employee →