// AI NEWS · MODEL RELEASE

DeepSeek Opens a Short Beta of V4.1 Flash, a New Multimodal Build It Says Surpasses V4 Pro

The limited test model runs under the id deepseek-v4.1-flash-expires-on-0910 and is priced like V4 Flash, with a native-multimodal architecture and an official release planned around September 10.

2026-09-09 · by Moe Ameen

Published 2026-09-09 · Updated

Short answer

DeepSeek V4.1 Flash is a natively multimodal Flash model that DeepSeek tested in a short beta under the id deepseek-v4.1-flash-expires-on-0910, at V4 Flash prices and up to 20 concurrent requests per account. It went live on September 10, 2026. DeepSeek claims it beats V4-Pro on performance, cost, speed and total time, but dropped its plan to route V4-Pro traffic to it.

Key takeaways

  • DeepSeek ran a short V4.1 Flash beta under the model id deepseek-v4.1-flash-expires-on-0910.
  • Beta pricing matched V4 Flash, with up to 20 concurrent requests per account.
  • DeepSeek says V4.1 Flash beats V4-Pro on performance, cost, speed and total time.
  • A plan to route V4-Pro requests to V4.1 Flash was reversed; V4-Pro service continues.

What is DeepSeek V4.1 Flash?

In early September 2026, DeepSeek quietly opened a limited-time beta of DeepSeek-V4.1-Flash, an interim test build it did not present as a formal release. Developers can call it through DeepSeek's existing API — the base URL is unchanged — by selecting the model id `deepseek-v4.1-flash-expires-on-0910`. As the name signals, the test build was scheduled to go offline on September 10, the day DeepSeek released V4.1 Flash officially. Beta pricing matches DeepSeek-V4-Flash, with each account limited to about 20 concurrent requests.

The headline change is the architecture. DeepSeek describes V4.1 Flash as a re-architected model that natively integrates multimodal capabilities, so it accepts images alongside text rather than routing vision through a separate experimental build like the earlier DeepSeek-V4-Flash-Vision-Exp. Alongside multimodality, DeepSeek claims stronger overall performance, faster generation, and lower cost.

The comparison DeepSeek is drawing is the notable part. The company says V4.1 Flash surpasses the more expensive DeepSeek-V4-Pro on performance, cost, speed, and total time. DeepSeek also announced a routing plan: from September 14, requests to V4-Pro would be served by V4.1 Flash and billed at the cheaper Flash rate until a V4.1 Pro arrived. It reversed that before it took effect, and V4-Pro service continues with billing unchanged.

As this is a short pre-release beta, treat the specifics — exact benchmarks, final pricing, and the release date — as provisional until DeepSeek publishes them in its changelog; confirm against the source before building anything permanent on the test model id, which expires by design.

Why does a cheaper multimodal Flash model matter for creators?

  • If a cheaper Flash tier really beats the flagship, the price of good drafting drops again. DeepSeek planned to move V4-Pro traffic onto V4.1 Flash automatically, then backed off, so switching is now a choice you make yourself.
  • Native multimodality folds image-reading into the mainline model, so a single cheap call can look at a screenshot, chart, or reference frame and reason over it in text — no separate vision endpoint.
  • The gain is upstream. A better, cheaper, image-aware writer still renders no video, cuts no clip, designs no carousel, and posts nothing — production and distribution stay the real work.
  • Don't build on the beta id. The test model expires by design (deepseek-v4.1-flash-expires-on-0910); wire anything durable to the official model name once DeepSeek ships it.
  • Model choice keeps commoditizing. With frontier-class, multimodal text this cheap, the durable advantage is the engine that turns that text into finished, on-brand, multi-platform content.

How do you turn V4.1 Flash drafts into finished posts with Kompozy?

There are two ways to act on this today, and both point the same direction. First, the practical one: if you already draft on DeepSeek, moving from V4-Pro to V4.1 Flash should make your scripts, captions, and outlines cheaper and, by DeepSeek's account, better. That widens the same bottleneck it always does — you can now afford to write far more than you can produce and publish. Kompozy is the engine that closes that gap. Draft or batch your copy on V4.1 Flash for pennies, bring it into Kompozy as source, and the engine makes the actual assets a model never will: a Persona Shorts avatar video reading your script, a brand-exact Carousel, Quote Graphics, Photo Posts, a Blog Article, and an Email Newsletter, all held to one voice by the Persona Brief, then scheduled across eight social platforms plus blog and email on autopilot.

Second, there's a fast play on the news itself. "DeepSeek V4.1 Flash" is a high-intent search this week while developers hunt for what changed and whether the V4 Pro claim holds. A single sharp point of view becomes a week of content inside Kompozy — a short explainer, a carousel breaking down the routing change, captioned clips, and platform-native posts — published while the interest curve is still climbing. DeepSeek is shipping a cheaper, smarter model; Kompozy is how that model's output turns into posts on every platform your audience actually uses. (Kompozy's own copy generation runs on Claude and OpenAI, so V4.1 Flash is your upstream drafting choice and Kompozy is the downstream engine.)

What are the key facts about DeepSeek V4.1 Flash?

  • DeepSeek opened a limited beta of V4.1 Flash in early September 2026, model id deepseek-v4.1-flash-expires-on-0910, expiring Sept 10, the day of the official release.
  • It is a re-architected, natively multimodal build priced the same as V4-Flash, with ~20 concurrent requests per account.
  • DeepSeek claims V4.1 Flash surpasses the larger V4-Pro on performance, cost, speed, and total time.
  • DeepSeek first planned to route V4-Pro requests to V4.1 Flash at the cheaper rate, then reversed course; V4-Pro service continues with billing unchanged.
  • Cheaper, image-aware drafting shifts the bottleneck to production and distribution — where Kompozy turns text into finished, published multi-platform content.

What does this look like in Kompozy?

A persona tweet graphic from Kompozy. Drafted text ends up as a finished, branded post like this. DeepSeek did not write it.

Persona tweet graphic generated in Kompozy
Real output · Kompozy Persona tweet format · from the Kompozy workspace

Who is this for, and who should skip it?

Best for:

  • Developers who draft text on DeepSeek and want a cheaper model that can also read images.
  • Teams comparing V4-Pro costs against a Flash tier.

Not ideal for:

  • Anyone building on the beta id. It expired by design on September 10.
  • People who need video, images or publishing. V4.1 Flash writes and reasons in text.

Frequently asked questions

What is DeepSeek V4.1 Flash?

It is a new, re-architected build of DeepSeek's fast Flash tier that natively supports multimodal input (images alongside text). DeepSeek opened a limited beta in early September 2026 under the model id deepseek-v4.1-flash-expires-on-0910 and released it officially on September 10. Beta pricing matches DeepSeek-V4-Flash.

Is DeepSeek V4.1 Flash better than V4 Pro?

DeepSeek claims so — it says V4.1 Flash beats V4-Pro on performance, cost, speed, and total time. It also planned to route V4-Pro API requests to V4.1 Flash at the cheaper rate, then reversed that plan. Because it is a short pre-release beta, treat the exact benchmarks as provisional until DeepSeek publishes them.

How much does DeepSeek V4.1 Flash cost?

During the beta it is priced the same as DeepSeek-V4-Flash, with no beta surcharge and roughly 20 concurrent requests per account. Check DeepSeek's pricing page for current rates on the released model.

Can DeepSeek V4.1 Flash make videos or social posts?

No. It is a multimodal text model — it reads images and writes and reasons in text, but it generates no images, video, or audio and publishes nothing. To turn its drafts into finished posts, pair it with a content engine like Kompozy, which generates video, carousels, and images and publishes across the eight social platforms plus blog and email.

How do we know this?

We checked every figure, date and quote on this page against the sources listed below on 2026-10-23, and removed or rewrote claims we could not confirm. Kompozy plan and format details come from kompozy.io/pricing.

Sources

  1. TechNode — DeepSeek opens limited beta of V4.1 Flash (read 2026-10-23)
  2. Yotta Labs — DeepSeek V4.1 Flash pricing, specs and V4 Pro routing (read 2026-10-23)
  3. Kompozy — Pricing (read 2026-10-23)

Related news

← All AI news · Get started →