---
id: "2026-04-24-deepseek-v4-preview"
url: "https://postcutoff.com/e/2026-04-24-deepseek-v4-preview/"
as_of: "2026-11-13T13:49:00+01:00"
date: "2026-04-24"
date_precision: day
category: model-release
importance: 5
confidence: high
status: [Confirmed]
sources: 4
editor: human editor
human_review: null
version: "2026-10-07"
---

As of: 2026-11-13 13:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/e/2026-04-24-deepseek-v4-preview/

# DeepSeek V4 preview: a 1.6T-parameter open MoE running on Huawei Ascend

Full title: DeepSeek V4 preview: 1.6T-parameter open MoE running on Huawei Ascend

DeepSeek released a preview of V4 on 2026-04-24: V4-Pro (1.6T total / 49B active) and V4-Flash (284B / 13B active), both MIT-licensed MoE models with a 1M-token context, validated on Huawei Ascend NPUs as well as Nvidia GPUs, priced far below Western frontier APIs.

## Key facts
- V4-Pro: 1.6T total parameters, 49B active; V4-Flash: 284B total, 13B active (The Register)
- Training data: 33T tokens; context window 1M tokens
- KV cache 9.5x-13.7x smaller than DeepSeek V3.2; mixed FP8/FP4 precision with quantization-aware training of MoE experts
- New hybrid attention (Compressed Sparse Attention + Heavily Compressed Attention) and Muon optimizer
- API price: Flash $0.14/M input, $0.28/M output; Pro $1.74/M input, $3.48/M output
- Day-zero support on Huawei Ascend SuperNode line incl. Ascend 950; weights on Hugging Face under MIT license

## What happened

On Friday 2026-04-24 DeepSeek published a **preview** of its fourth-generation model family. Two MoE models shipped:
**V4-Pro** (1.6 trillion parameters, 49B active) and **V4-Flash** (284B, 13B active), both with a 1M-token context window and trained on ~33T tokens.
Architecturally, DeepSeek introduced a hybrid compressed attention scheme and adopted the Muon optimizer, and cut KV-cache memory 9.5-13.7x versus V3.2,
using FP8/FP4 mixed precision with quantization-aware training.

The launch was notable for hardware: DeepSeek validated the models on **Huawei Ascend** NPUs (Huawei announced day-zero support across its SuperNode line,
including Ascend 950) as well as Nvidia GPUs. Coverage (Tom's Hardware) linked the release to escalating US government accusations of IP theft / distillation by Chinese labs.
Later milestones: V4-Flash re-post-trained update (2026-07-31), V4-Pro GA with low/high/max thinking effort (2026-08-13), and V4.1-Flash (2026-09-10).

## Why it matters

V4 was the largest open-weights model at release and the first frontier-class release optimized for a Chinese AI accelerator, a signal that China's
model stack can decouple from Nvidia. Its aggressive pricing (Pro output $3.48/M) kept pressure on Western API prices.

## Your AI and this story

- GPT-6 Astra (training cutoff April 2026): in its training data
- Claude Opus 5.5 (training cutoff June 2026): in its training data
- Gemini 4 Hydrogen (training cutoff March 2026): 24 days after its cutoff
- Grok 4.7 (training cutoff May 2026): in its training data

## Sources

1. [DeepSeek API changelog](https://api-docs.deepseek.com/updates/) (api-docs.deepseek.com, docs)
2. [The Register: DeepSeek's new models offer big inference cost savings](https://www.theregister.com/2026/04/24/deepseek_v4/) (theregister.com, press)
3. [Tom's Hardware: DeepSeek launches 1.6T V4 on Huawei chips](https://www.tomshardware.com/tech-industry/artificial-intelligence/deepseek-launches-1-6-trillion-parameter-v4-on-huawei-chips-as-us-escalates-ai-theft-accusations) (tomshardware.com, press)
4. [Huawei Central: DeepSeek launches V4 on Huawei chips](https://www.huaweicentral.com/deepseek-launches-new-v4-ai-models-running-on-huawei-chips/) (huaweicentral.com, press)

## Changes

- 2026-09-29 (filed): Created

## Related

- 2026-09-18: [Huawei sets Ascend 950 cluster cloud launch and Ascend 960 roadmap](https://postcutoff.com/e/2026-09-18-huawei-ascend-950-cluster-cloud/index.md)
- 2026-09-10: [DeepSeek V4.1-Flash brings a new architecture family, native vision and a cheaper API](https://postcutoff.com/e/2026-09-10-deepseek-v4-1-flash/index.md)