Skip to content
CheapestInference

DeepSeek V4 Flash API — retired (upgraded to DeepSeek V4.1 Flash)

DeepSeek V4 Flash was DeepSeek’s fast model. It was served in the Core Pool until September 10, 2026, when the pool’s DeepSeek slot was upgraded in place to DeepSeek V4.1 Flash — a new generation with native image input and a 1M-token context, at the same flat price. Core subscriptions were not interrupted and nothing changed on existing keys.

What this means

Section titled “What this means”
  • Requests with model: "deepseek-v4-flash" are served by DeepSeek V4.1 Flash until October 10, 2026. After that date the id returns an invalid-model error.
  • Core Pool subscriptions are unchanged: DeepSeek V4.1 Flash and MiMo v2.5 remain served with no token caps.
  • GET /v1/models is always the authoritative live list.

Migrate in one line

Section titled “Migrate in one line”

Change the model id — the endpoint, key, and subscription stay the same:

from openai import OpenAI


client = OpenAI(
    base_url="https://api.cheapestinference.com/v1",
    api_key="sk-..."  # your subscriber key
)


response = client.chat.completions.create(
    model="deepseek-v4.1-flash",  # was "deepseek-v4-flash"
    messages=[{"role": "user", "content": "Summarize this document..."}],
)
Terminal window
curl https://api.cheapestinference.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "deepseek-v4.1-flash", "messages": [{"role": "user", "content": "Hello"}]}'

Coding agents: set the model to deepseek-v4.1-flash in your tool’s config — see DeepSeek with Claude Code, Cline / Roo Code and OpenCode, Roo & Aider.

DeepSeek V4 Flash in brief (historical)

Section titled “DeepSeek V4 Flash in brief (historical)”
Model IDdeepseek-v4-flash (retired — use deepseek-v4.1-flash)
ServedCore Pool, until 2026-09-10
Last buildV4-Flash-0731 — agent-focused retrain, live from 2026-07-31
SuccessorDeepSeek V4.1 Flash — 552B MoE, native image input, 1M context, MIT weights

Common questions

Section titled “Common questions”

Is there still a DeepSeek V4 Flash API? Not as a separate model. Requests for deepseek-v4-flash are served by DeepSeek V4.1 Flash until October 10, 2026; use deepseek-v4.1-flash going forward.

Do I lose anything with the upgrade? No — DeepSeek V4.1 Flash is a newer generation at the same price and the same 1M-token context as served here, and it adds image input. The upgrade was applied to every Core subscription automatically.

Do I need a new subscription or key? No. The same Core Pool subscription and API key serve DeepSeek V4.1 Flash and MiMo v2.5 — just change the model id.

Section titled “Related”