Model · DeepSeek
DeepSeek V4 Pro: price, limits and messages on every plan
DeepSeek's DeepSeek V4 Pro in llmwise: how many DeepSeek V4 Pro messages each plan gives, what counts against them, what DeepSeek V4 Pro reads, where your messages go, and what it costs at DeepSeek's API prices.
Prices and limits checked against OpenRouter's DeepSeek V4 Pro page. Updated .
Short answer
DeepSeek V4 Pro isn't in llmwise's free trial; it needs a paid plan. A paid plan gives up to 125 DeepSeek V4 Pro messages a month on Plus ($10), 250 on Pro ($20), 800 on Max ($50), 1,800 on Ultra ($100) and 4,000 on Studio ($200). Each DeepSeek V4 Pro message's price is shown before you send it.
DeepSeek V4 Pro on every plan
What DeepSeek V4 Pro costs in llmwise is a count, not a token bill: up to 250 a month on Pro, and it isn't in the free trial.
| Plan | Price | DeepSeek V4 Pro messages | Each, if all go to DeepSeek V4 Pro |
|---|---|---|---|
| Free | $0 | Not in the trial | — |
| Plus | $10 a month | Up to 125 a month | $0.08 |
| Pro | $20 a month | Up to 250 a month | $0.08 |
| Max | $50 a month | Up to 800 a month | $0.06 |
| Ultra | $100 a month | Up to 1,800 a month | $0.06 |
| Studio | $200 a month | Up to 4,000 a month | $0.05 |
Prices don't include tax, which is added where it applies and shown before you pay. A paid plan's month is one allowance every model shares, so each DeepSeek V4 Pro count is the most you get if all of it goes to DeepSeek V4 Pro, and "each" is the plan's price divided by that count. How pricing works.
Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Plus $3.75, Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.
DeepSeek V4 Pro usage limits, explained
Every limit on DeepSeek V4 Pro in llmwise is published. Here's each one, in DeepSeek V4 Pro messages.
DeepSeek V4 Pro draws on the monthly allowance every model shares: 250 DeepSeek V4 Pro messages on Pro if nothing else uses it. Each message on a smaller model leaves fewer DeepSeek V4 Pro messages, each DeepSeek V4 Pro message leaves fewer for them, and unused DeepSeek V4 Pro messages don't carry over when the billing period renews.
Long DeepSeek V4 Pro chats count more, because each DeepSeek V4 Pro reply resends the whole conversation. Past 64k tokens a DeepSeek V4 Pro reply counts as 2 messages, so Pro's 250 DeepSeek V4 Pro messages become up to 125 such replies; past 128k it counts as 4 (up to 62 DeepSeek V4 Pro replies); past 200k the DeepSeek V4 Pro chat is full.
DeepSeek V4 Pro's native context window is 1.05M tokens; llmwise stops a DeepSeek V4 Pro chat at 200k tokens, so a DeepSeek V4 Pro chat uses at most 200k of its 1.05M tokens.
A DeepSeek V4 Pro reply can run to 8,000 tokens on Plus and Pro and 16,000 on Max, Ultra, and Studio, at the same DeepSeek V4 Pro price however long it runs.
On a paid plan each web search or link read in a DeepSeek V4 Pro reply counts as 1 more DeepSeek V4 Pro message.
A $10 top-up adds up to 200 more DeepSeek V4 Pro messages, valid 12 months, spent once the plan's own DeepSeek V4 Pro count is gone.
DeepSeek V4 Pro replies share every plan's soft cap of 100 replies an hour with the other models.
What DeepSeek V4 Pro reads and does
| Made by | DeepSeek |
|---|---|
| Model at OpenRouter | deepseek/deepseek-v4-pro-0813 |
| Context window | 1.05M tokens |
| Longest chat in llmwise | 200k tokens |
| Images you attach | No |
| PDFs you attach | Gets the text only |
| Tools (search, documents, code, apps) | Yes |
| Reasoning | On, at medium effort |
| API price, per million tokens | $0.40 in, $4.00 out |
| A typical message at API prices | $0.0044 |
| On Pro | Up to 250 a month |
Where your DeepSeek V4 Pro messages go
llmwise doesn't call DeepSeek for DeepSeek V4 Pro: every DeepSeek V4 Pro message goes through OpenRouter, which routes it to a company hosting DeepSeek V4 Pro.
Each DeepSeek V4 Pro request asks OpenRouter only for DeepSeek V4 Pro hosts that, according to OpenRouter, don't store prompts for their own use or train on them, and skips any listed as running a 4-bit (fp4) build of DeepSeek V4 Pro. The DeepSeek V4 Pro host that answers can be a hosting company rather than DeepSeek, under its own terms.
If one DeepSeek V4 Pro host is down, OpenRouter moves the DeepSeek V4 Pro request to another that meets the same terms.
The Privacy Policy lists every company a DeepSeek V4 Pro message can reach.
DeepSeek V4 Pro's API price, and what a message costs
DeepSeek V4 Pro's API price is $0.40 per million tokens in and $4.00 out: the price of the OpenRouter endpoints llmwise uses, not DeepSeek's own API. At that price a typical DeepSeek V4 Pro message (4,000 tokens in, 700 out) costs about $0.0044.
In llmwise a DeepSeek V4 Pro message isn't billed by the token: it's one of up to 250 DeepSeek V4 Pro messages a month on Pro, however long the DeepSeek V4 Pro reply runs.
What DeepSeek V4 Pro's replies cost in our test runs, job by job, and DeepSeek V4 Pro beside the rest of its family: DeepSeek V4 Pro pricing: API price, cost per message and plans.
When to use DeepSeek V4 Pro, and when a cheaper model will do
llmwise's catalog describes DeepSeek V4 Pro as “DeepSeek's most capable model”. No benchmark scores here: ask DeepSeek V4 Pro your own question, then switch the picker to another model in the same chat, which sees DeepSeek V4 Pro's answer too.
When a question doesn't need DeepSeek V4 Pro, these read the same attachments and leave more of the month for DeepSeek V4 Pro:
DeepSeek V4.1 Flash (DeepSeek): 60 a day on Pro, against DeepSeek V4 Pro's 250.
Claude Haiku 5.5 (Anthropic): 60 a day on Pro, against DeepSeek V4 Pro's 250.
GPT-6 Luna (OpenAI): 60 a day on Pro, against DeepSeek V4 Pro's 250.
GLM 5.3 Flash (Z.ai): 60 a day on Pro, against DeepSeek V4 Pro's 250.
Same count as DeepSeek V4 Pro on every plan: Claude Haiku 4.5, Gemini 3.8 Flash, Grok 4.7, GLM 5.3 and Mistral Large 4.
DeepSeek V4 Pro in our test runs
DeepSeek V4 Pro passed 49 of our 50 test prompts, sent through llmwise's own pipeline and checked the same way for every model.
- Coding5 of 5 passed
- Writing5 of 5 passed
- Math5 of 5 passed
- Summarization5 of 5 passed
- Data analysis5 of 5 passed
- Customer support4 of 5 passed
- Translation5 of 5 passed
- SQL5 of 5 passed
- RAG and answering from documents5 of 5 passed
- Agents and tool use5 of 5 passed
How DeepSeek V4 Pro began its reply to our first writing prompt, “Write a LinkedIn post of 100 to 150 words announcing that our small bakery, Crumb & Co., opens a second shop in Leeds on 14 November. Mention the free coffee…”:
Big news from Crumb & Co. — we're opening a second shop, and this time we're heading to Leeds. Our new bakery will open its doors on Thursday 14 November, and we'd love you to be among the first through them. Come and see the new space, smell the bread coming out of the oven, and say hello to the team. To mark the day, we're giving away free coffee to everyone who visits — our way of saying thank you for supporting a small, independent bakery. We've spent months getting this little shop ready, and we're equal parts nervous and excited. If you're in Leeds, we'd be glad to see you. Bring a friend, grab a warm loaf, and help us make it feel like home. See you on the 14th.
Also about DeepSeek V4 Pro: DeepSeek V4, without a DeepSeek account
Questions
Is DeepSeek V4 Pro free in llmwise?
No. DeepSeek V4 Pro isn't in llmwise's free trial, which is Claude Haiku 5.5 alone. Pro ($20 a month) gets up to 250 DeepSeek V4 Pro messages a month.
What are DeepSeek V4 Pro's usage limits in llmwise?
DeepSeek V4 Pro has a published monthly count on each paid plan (up to 250 DeepSeek V4 Pro messages a month on Pro), shared with the other models. A long DeepSeek V4 Pro chat counts each reply as 2 DeepSeek V4 Pro messages past 64k tokens and 4 past 128k.
What happens when my DeepSeek V4 Pro messages run out?
Your DeepSeek V4 Pro count resets when your billing period renews. Until then, the everyday models keep answering on their own daily count, and a $10 top-up adds up to 200 more DeepSeek V4 Pro messages.
How much context does DeepSeek V4 Pro have in llmwise?
1.05M tokens is DeepSeek V4 Pro's native context window. A DeepSeek V4 Pro chat in llmwise stops at 200k tokens, and past 64k each DeepSeek V4 Pro reply counts as more than one message, so a new DeepSeek V4 Pro chat for each new task goes further.
Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral, in one chat.
See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.