Is Qwen better than DeepSeek?
Neither is ranked above the other, because Qwen and DeepSeek are shaped differently. DeepSeek ships one generation, V4, under MIT for both code and weights, with a native 1 million token context and the lowest published per-token price on this page. Qwen ships a deeper Apache 2.0 ladder from 0.8B to 397B, a multimodal open model, and a dedicated coding line, next to a separate API-only line. Pick DeepSeek for license clarity, long-context text work, and cost per token. Pick Qwen for a model size that fits your hardware, or for vision in an open model.
What TSK-1 Found
TSK-1 has no result table that puts these two families side by side yet; here's what each record shows. Qwen entered TSK-1 on Aug 25, 2026: Qwen 3.7 Plus finished a 32-question client sign-up form with every answer saved, and Qwen 3.7 Max built a clean match tracker on its first run. DeepSeek V4 has been tested several times. DeepSeek V4 Flash built the best-looking app of its test (Aug 1, 2026), and on Aug 3, 2026 it was the cheapest, the fastest, and the cleanest all at once. TSK-1 tested Qwen's API models, not its open weights, so that side of the comparison rests on published vendor benchmarks and license terms.
- DeepSeek: Aug 1, 2026 — best-looking app of its test. Aug 3, 2026 — cheapest, fastest, and cleanest all at once. Aug 6, 2026 — best at holding your information (60 separate fields captured).
- Qwen: Tested in TSK-1 on Aug 25, 2026, scoring 50 out of 100 (dated findings).
See the full evidence at /tsk/deepseek, /tsk/qwen, and the TSK-1 hub.
The Distinction That Decides Everything
Qwen's open weights and Qwen's API tiers are different generations of model. Every tier whose name ends in -max, -plus, or -flash is proprietary and API-only, and the open ladder sits a generation behind it. DeepSeek does not work that way: the models it runs are the models it publishes, all under MIT.
Get that wrong and everything downstream is wrong — the license you think you have, the model you think you can self-host, and the price you think you are paying.
TL;DR: Qwen publishes Apache 2.0 weights at Qwen3.6 (a 27B dense model and a multimodal 35B-A3B mixture-of-experts) while selling a separate API-only line that differs between international and China mainland endpoints. DeepSeek publishes one generation, V4, under MIT, at $0.44 / $1.32 per million tokens on Flash at peak. Route both inside Taskade Genesis.
What Ships Open, and What Only Ships as an API
Open, with downloadable weights under Apache 2.0:
- Qwen3.6-27B — dense.
- Qwen3.6-35B-A3B — Mixture-of-Experts, 35B total and 3B used per token, and multimodal via a vision encoder that handles both images and video.
- The Qwen3.5 ladder — 397B-A17B, 122B-A10B, 35B-A3B, 27B, 9B, 4B, 2B, and 0.8B, which is an unusually complete range of sizes.
Proprietary, API-only: qwen3.7-max, qwen3.7-plus, qwen3.6-flash, qwen3.5-omni-plus, and every other -max / -plus / -flash tier. No Qwen3.7 weights exist in the official Hugging Face organization, so do not describe those tiers as open.
Alibaba's own wording on the open half is unambiguous: "All our open-weight models are licensed under Apache 2.0."
Qwen 3.7 Max vs DeepSeek V4 Pro
Qwen 3.7 Max and DeepSeek V4 Pro are the top API tiers of each family, and they are built on opposite terms. qwen3.7-max is proprietary and API-only: no Qwen3.7 weights exist in the official Hugging Face organization, and on the international Model Studio it lists at $2.50 per million input tokens and $7.50 per million output. China mainland runs a generation ahead at the top, with qwen3.8-max. deepseek-v4-pro is MIT for code and weights, 1.6T total parameters with 49B used per token, a native 1M-token context with a 384K maximum output, at $1.32 in and $3.96 out at peak and exactly half that off-peak.
Both built the same match tracker in TSK-1, in different tests. Qwen 3.7 Max built a complete one with a clean dashboard on its first run (Aug 25, 2026), though with no hero picked its save button did nothing and said nothing. DeepSeek V4 Pro built one of the two best-looking trackers of its test (Sep 22, 2026), but its hero picker listed only 17 heroes. Pick qwen3.7-max when you want the top of Alibaba's commercial line, the broader multilingual bet, on the international endpoint. Pick deepseek-v4-pro for open weights, a native million-token window, and the lower published rate.
Region Changes the Model List
There is no single global Qwen lineup, and quoting one is the fastest way to send a reader to a model they cannot call.
| Endpoint | Top tier | Mid tier | Fast tier |
|---|---|---|---|
| International (Singapore) Model Studio | qwen3.7-max |
qwen3.7-plus |
qwen3.6-flash |
| China mainland | qwen3.8-max |
qwen3.7-plus |
qwen3.7-flash |
Mainland runs ahead of international, and mainland token prices are lower. Always name the region alongside the model.
What Runs on the DeepSeek Side in August 2026
DeepSeek moved a whole generation this summer, and a lot of published guidance is now describing endpoints that no longer answer.
- V3 and R1 are gone, and there is no R2. The current generation is V4.
deepseek-chatanddeepseek-reasonerwere retired and became inaccessible on 24 July 2026.- The two production ids are
deepseek-v4-flashanddeepseek-v4-pro, plus an experimentaldeepseek-v4-flash-vision-expon Flash's rate card. - Reasoning is a mode, not a model.
deepseek-v4-flashexposes reasoning effort atlow,high, andmax. - Parameters:
deepseek-v4-prois 1.6T total with 49B used per token.deepseek-v4-flashis roughly 284–304B total with 13B used per token — vendor sources disagree on the total, so treat it as a range.
Licenses: Apache 2.0 vs MIT
| Qwen open weights | DeepSeek V4 | |
|---|---|---|
| License | Apache 2.0, no size exceptions | MIT |
| Covers the weights | ✅ Yes | ✅ Yes |
| Covers the repository code | ✅ Yes | ✅ Yes, explicitly |
| Commercial use | ✅ Permitted | ✅ Permitted |
| Redistribute fine-tunes | ✅ Permitted with attribution | ✅ Permitted, retain notice and state changes |
| Monthly-active-user cap | None | None |
| Applies to every model in the family | 🔴 No — only the open ladder | ✅ Yes |
Both are genuinely permissive, and both are a different universe from a closed frontier model, where self-hosting and on-prem data residency are simply not purchasable at any price. The practical difference is coverage: MIT covers everything DeepSeek ships, while Apache 2.0 covers only the half of Qwen that ships as weights.
Benchmarks: What We Can Actually Cite
The published benchmark for Qwen3.6-35B-A3B is SWE-bench Verified 73.4% and MMLU-Pro 85.2% — notable because that is a 35B-total model activating 3B parameters per token, not a frontier-scale one.
We are deliberately not printing a benchmark table pitting the two families against each other. Scores move between point releases, vendors publish under different harnesses, and a comparison page that quotes a stale leaderboard does more harm than an honest gap. Run your own evaluation on your own prompts before you commit a pipeline to either.
Price per Million Tokens
International Model Studio for Qwen (mainland is lower), published DeepSeek API rates for V4.
| Model | Input | Output |
|---|---|---|
deepseek-v4-flash — peak |
$0.44 (cache hit $0.014) | $1.32 |
deepseek-v4-flash — off-peak |
$0.22 (cache hit $0.007) | $0.66 |
deepseek-v4-pro — peak |
$1.32 (cache hit $0.044) | $3.96 |
deepseek-v4-pro — off-peak |
$0.66 (cache hit $0.022) | $1.98 |
qwen3.6-flash (intl) |
$0.25 up to 256K, $1.00 from 256K–1M | $1.50 / $4.00 |
qwen3.7-plus (intl) |
$0.40 up to 256K, $1.20 from 256K–1M | $1.60 / $4.80 |
qwen3.7-max (intl) |
$2.50 | $7.50 |
Two things to hold on to. First, Qwen's commercial tiers price in context bands — cross 256K and the input rate roughly triples on qwen3.6-flash, which changes the economics of long-document work more than the headline number suggests. Second, DeepSeek prices on a clock — off-peak is exactly half of peak, and peak covers only 01:00 to 04:00 and 06:00 to 10:00 UTC on weekdays. Batch work you can schedule outside those windows costs half without changing a line of code. Every figure in this table is the rate published as of August 2026.
A third point worth its own line: on DeepSeek a cache hit is roughly 31× below a cache miss on flash and exactly 30× below on pro, at either clock rate, which is a far bigger lever than choosing between the two models. Qwen's cached-input prices for the commercial tiers are not published in a form we can cite, so no equivalent figure appears for them above.
The Routing Matrix
Pick per task, and keep the cost consequence in the same view as the capability.
| Task | Reach for | Why | Cost consequence |
|---|---|---|---|
| Cheap high-volume classification | deepseek-v4-flash |
Lowest published rate of the two families | $0.44 in / $1.32 out per 1M at peak, $0.22 / $0.66 off-peak |
| Repeated prompts over a fixed corpus | deepseek-v4-flash with cache hits |
Stable prefix bills at $0.014 per 1M at peak, $0.007 off-peak | Roughly 31x off the input side |
| Long-context reasoning, 1M tokens | deepseek-v4-pro |
1M native context, no scaling technique required | $1.32 / $3.96 per 1M at peak — a 500K prompt is a real line item |
| Vision — image or video in the loop | Qwen3.6-35B-A3B |
The multimodal model that also ships as open weights | Self-host cost, or your provider's rate |
| Agentic coding | Qwen3-Coder variants, or V4 at high effort |
Qwen keeps a dedicated coder line; DeepSeek uses one endpoint with a dial | Coder-Next ~80B is far cheaper to run than 480B-A35B |
| Multilingual content | Qwen commercial tiers, region-matched | Alibaba's commercial line is the broader multilingual bet | qwen3.7-plus $0.40 / $1.60 up to 256K, intl |
| On-prem, air-gap, data residency | Qwen3.6-35B-A3B or deepseek-v4-flash |
Apache 2.0 and MIT both permit self-hosting outright | Your GPU cost, no per-token bill |
| Smallest viable footprint | Qwen3.5 ladder down to 0.8B | Nothing else here goes that small | Runs where a frontier model cannot |
| Overnight or weekend batch runs | Either DeepSeek model, scheduled off-peak | DeepSeek halves the rate outside peak hours | 50% off input and output |
| A workflow that needs several of these | Route across models | No single family wins every row | Allowance included in the subscription |
Self-Host Reality Check
Both licenses permit self-hosting. Hardware is what decides whether you actually will.
Qwen3.6-35B-A3Bis the realistic target. 35B total with only 3B used per token is the reason Qwen's open ladder shows up so often in on-prem deployments, and it is multimodal on top of that.- The Qwen3.5 ladder gives you an exit at every size — 397B-A17B at the top, 0.8B at the bottom. If a model does not fit your hardware, there is a smaller one with the same license.
deepseek-v4-proat 1.6T total is a serious commitment.deepseek-v4-flashat roughly 284–304B total and 13B used per token is the more approachable DeepSeek self-host, but it is still an order of magnitude above the practical Qwen target.
For most teams below serious scale, a managed gateway is cheaper and simpler than running either. The license still matters even if you never self-host, because it is what makes the option available at all when a compliance requirement lands.
Choose DeepSeek If
A comparison that only flatters one side is not worth reading. DeepSeek is the better pick when:
- You want one license covering everything you touch. MIT applies to the code and the weights, and the models DeepSeek runs are the models it publishes. There is no open-versus-commercial line to keep track of.
- You need a million tokens of context natively. Both V4 models carry it without a scaling technique, with a 384K maximum output.
- Per-token cost is the binding constraint. On published international rates,
deepseek-v4-flashis the cheapest option on this page by a wide margin, and its cache-hit rate, at $0.014 per 1M at peak, is roughly 31x lower still.
Choose Qwen when you need vision in an open model, a specific parameter size that fits your hardware, a dedicated coding line, or a commercial multilingual tier — and when you are comfortable tracking which half of the family a given model belongs to.
The Taskade Angle: Route Across Labs, Not Between Them
Taskade runs frontier models from top AI labs inside a single workspace, with the AI allowance included in the subscription instead of an API account and a key per lab. Plans are Free, Pro $10/mo billed annually, Business $25/mo billed annually, Max $100/mo billed annually, and Enterprise $250/mo billed annually.
What that changes day to day:
- Per-step model choice. Each agent or automation step picks its own model, so bulk extraction and final reasoning do not have to share a bill.
- No account sprawl. One workspace instead of a separate console, key, and invoice for every lab.
- Routing survives a repricing or a retirement. DeepSeek retired two model ids outright this summer and has announced a price rise; that is a routing change, not a migration.
- Work that outlives the prompt. Projects, agents, and automations sit alongside 100+ integrations that flow both ways — triggers pull events in from tools like Slack, Gmail, and Google Calendar, and actions push data out to Stripe, Notion, and GitHub.
Model availability is flag-driven and changes as labs ship, so treat the in-product picker rather than any comparison page as the source of truth for what you can call today.
See 10 Best Open-Source AI LLMs in 2026 for the wider open-weight field.
Frequently Asked Questions
Which is better, Qwen or DeepSeek?
They are shaped differently rather than ranked. Qwen ships a wide ladder of Apache 2.0 open-weight models alongside a separate proprietary API line, and one of its open models is multimodal. DeepSeek ships a single generation, V4, published under MIT for both the code and the weights, in two sizes with a 1 million token context. If your constraint is license clarity and long-context text work, DeepSeek is the simpler story. If your constraint is picking a model size that fits your hardware, or you need vision in an open model, Qwen's ladder is the deeper bench.
Are Qwen and DeepSeek open source under the same license?
No, and the difference is bigger than the license names suggest. Qwen states that all of its open-weight models are licensed under Apache 2.0, with no size exceptions, which is genuinely permissive. But only some Qwen models are open at all. DeepSeek publishes under MIT, and MIT covers both the repository code and the model weights, which is the cleanest commercial position of the two. Both permit commercial use, fine-tuning, and redistribution with attribution.
Which Qwen models actually have downloadable weights?
The Qwen3.6 open models are Qwen3.6-27B, a dense model, and Qwen3.6-35B-A3B, a Mixture-of-Experts model with 35 billion total parameters and 3 billion used per token that is also multimodal with a vision encoder handling images and video. The Qwen3.5 open line covers 397B-A17B, 122B-A10B, 35B-A3B, 27B, 9B, 4B, 2B, and 0.8B. Every tier ending in -max, -plus, or -flash is API-only, and no Qwen3.7 weights exist in the official Hugging Face organization.
Does the Qwen model list depend on which region you use?
Yes, and this trips up most comparisons. The international Model Studio, served from Singapore, offers qwen3.7-max, qwen3.7-plus, and qwen3.6-flash. China mainland runs ahead of it with qwen3.8-max, qwen3.7-plus, and qwen3.7-flash. There is no single global Qwen model list, so always check the endpoint you are actually calling before you assume a model name is available to you.
Which DeepSeek models are live right now?
Two: deepseek-v4-flash and deepseek-v4-pro. V3 and R1 are gone, there is no R2, and the deepseek-chat and deepseek-reasoner API ids were retired and became inaccessible on 24 July 2026. Reasoning is no longer a separate model either. Thinking is a mode inside V4, and deepseek-v4-flash exposes a reasoning-effort setting with low, high, and max levels.
How do Qwen and DeepSeek compare on context length?
The Qwen open-weight models carry a native context of 262,144 tokens, extensible to roughly 1,010,000 tokens using YaRN scaling. Both DeepSeek V4 models carry a 1 million token context natively, with a maximum output of 384,000 tokens. So DeepSeek gets you to a million tokens without a scaling technique, while Qwen gets there with one. Context windows for Qwen's commercial -max, -plus, and -flash tiers are best read off the region-specific documentation for the endpoint you use.
Which one is better for coding?
Qwen's coding line is still Qwen3-Coder. There is no Qwen3.5-Coder or Qwen3.6-Coder, which is a common mistake in guides that assume the coder models track the main generation. The variants are Qwen3-Coder-Next at roughly 80 billion parameters in a hybrid design, plus 480B-A35B and 30B-A3B, with 256K native context extensible to 1M via YaRN. On the DeepSeek side there is no separate coding model at all, because the V4 endpoints handle code with the same reasoning-effort dial as everything else.
Which is cheaper to run?
On published international rates, DeepSeek is cheaper per token, and it bills on a two-rate clock. At peak, deepseek-v4-flash is $0.44 per million input and $1.32 per million output, and deepseek-v4-pro is $1.32 and $3.96. Off-peak is exactly half of each: $0.22 / $0.66 on flash and $0.66 / $1.98 on pro. Peak covers 01:00 to 04:00 and 06:00 to 10:00 UTC, Monday to Friday, so most hours of the week bill at the lower rate. On Alibaba's international Model Studio, qwen3.6-flash is $0.25 per million input up to 256K context and $1.50 output, qwen3.7-plus is $0.40 and $1.60 in the same band, and qwen3.7-max is $2.50 and $7.50. China mainland rates are lower than the international ones. All of these are the rates published as of August 2026.
Should I self-host Qwen or DeepSeek?
It depends far more on model size than on brand. Qwen3.6-35B-A3B uses only 3 billion parameters per token, which makes it the most realistic self-host target of the two families and the reason Qwen's ladder is popular with teams running their own hardware. DeepSeek's deepseek-v4-pro is 1.6 trillion total parameters with 49 billion used per token, which is a serious infrastructure commitment, and deepseek-v4-flash sits at roughly 284 to 304 billion total with 13 billion used per token. Both licenses permit self-hosting outright, so the question is purely one of hardware and operational appetite.
Can I use Qwen and DeepSeek together inside Taskade?
Yes. Taskade runs frontier models from top AI labs inside a single workspace, with the AI allowance included in the subscription rather than a separate API account per lab. You can pick a different model per agent or per automation step, so a bulk extraction step and a final reasoning step do not have to share the same bill. Model availability is flag-driven and changes as labs ship, so the in-product picker is the source of truth for what is available today.
Final Word: Two Different Shapes of Open
Qwen gives you the deeper bench — an Apache 2.0 ladder from 0.8B to 397B, a multimodal open model, and a dedicated coding line — at the cost of tracking which models are actually open and which region you are calling. DeepSeek gives you one MIT-licensed generation where the published model and the served model are the same thing, with a native million-token context and the lowest published per-token price here.
Neither answer holds for every step of a real workflow. Route per task, and keep the price of each choice in front of you when you make it.
▲ Memory feeds Intelligence. ■ Intelligence triggers Execution. ● Execution creates Memory. Two open-weight families. One workspace.
This is the origin of living software. 🌱
Build with Qwen and DeepSeek in one workspace →
Related reading
- 10 Best Open-Source AI LLMs in 2026 — The full open-weight ranking.
- DeepSeek vs ChatGPT — Token pricing against a per-seat consumer product.
- Kimi vs DeepSeek — The other open-weight head-to-head.
- Multi-Model AI Access — How Taskade routes frontier models from top AI labs.
- Model Credits — Per-model credit costs and plan quotas.
- Taskade MCP Server — Use Claude Desktop or Cursor with your workspace.
- Free Claude Alternative — How premium frontier compares.
- Free ChatGPT Alternative — The OpenAI side.
- TSK-1 DeepSeek profile — Full benchmark evidence for the DeepSeek family.
- TSK-1 Qwen profile — Full benchmark evidence for the Qwen family.
- TSK-1 hub — The complete model benchmark dataset.
