2026-04-29原文

DeepSeek-V4 Pro now available on Together AI

DeepSeek-V4 Pro, a 1.6T-parameter MoE reasoning model, is now available on Together AI with a 512K context window, controllable reasoning modes, and cached-input pricing for long-context workloads like code agents, document intelligence, and research synthesis.

Article intelligence

EngineersAdvanced

Key points

1.6T-parameter MoE with 49B activated parameters, 512K context on Together AI (model supports 1M)
Three reasoning modes: Non-Think, Think High, Think Max to match effort to task
Pricing: $2.10/1M input, $0.20/1M cached input (90% savings), $4.40/1M output
Best suited for code agents, document intelligence, long-context agents, and research synthesis

Why it matters

This matters because 1.6T-parameter MoE with 49B activated parameters, 512K context on Together AI (model supports 1M).

Technical impact

May affect model selection, inference cost, product capability, and evaluation benchmarks.

DeepSeek-V4 Pro now available on Together AI

⚡️ FlashAttention-4: up to 1.3× faster than cuDNN on NVIDIA Blackwell →

Introducing Together AI's new look →

🔎 ATLAS: runtime-learning accelerators delivering up to 4x faster LLM inference →

⚡ Together GPU Clusters: self-service NVIDIA GPUs, now generally available →

📦 Batch Inference API: Process billions of tokens at 50% lower cost for most models →

🪛 Fine-Tuning Platform Upgrades: Larger Models, Longer Contexts →

Inference

Serverless Inference

High-performance inference as APIs

Batch Inference

Inference for batch workloads

Dedicated Model Inference

Inference on custom hardware

Dedicated Container Inference

Inference for custom models

MiniMax M2.5

Nano Banana Pro

Qwen3.5-397B

GLM-5

kimi k2.5

gpt-oss-120B

Model library

Explore the top open-source models

Compute

Accelerated Compute

GPU Clusters

Reliable GPU clusters at scale

AI Factory

Custom infrastructure at frontier scale

Developer Environments

Sandbox

Build development environments for AI

Storage

Managed Storage

Store model weights & data securely

GB300

GB200

B200

H200

H100

Model Shaping

Fine-Tuning

Shape models with your data

Evaluations

Measure model quality

DeepSeek V3.1

GLM 5 FP4

Qwen3-VL 32B

gpt-oss-120b

kimi k2.5

Llama 4 Maverick

Model library

Fine-tune top open-source models

Research

Systems research for production AI

Research blog

All our research publications

Featured publications

FlashAttention

ATLAS

Kernel Collection

ThunderKittens

DSGym

Show all

Developers

Documentation

Technical docs for Together AI

Demos

Our open-source demo apps

Cookbooks

Practical implementation guides

Voice Agents

Build voice agents for production

Model Library

Playground

Together Chat

Which LLM to use

Company

Resources

Customer stories

Testimonials from AI Natives

Startup accelerator

Build and scale your startup

Customer support

Find answers to your questions

Blog

Our latest news & blog posts

Events

Explore our events calendar

Company

About

Get to know us

Careers

Join our mission

Pricing

Serverless Inference

High-performance inference as APIs

Batch Inference

Inference for batch workloads

Dedicated Model Inference

Inference on custom hardware

Dedicated Container Inference

Inference for custom models

MiniMax M2.5

Nano Banana Pro

Qwen3.5-397B

GLM-5

kimi k2.5

gpt-oss-120B

Model library

Explore the top open-source models

Accelerated Compute

GPU Clusters

Reliable GPU clusters at scale

AI Factory

Custom infrastructure at frontier scale

Developer Environments

Sandbox

Build development environments for AI

Storage

Managed Storage

Store model weights & data securely

GB300

GB200

B200

H200

H100

Fine-Tuning

Shape models with your data

Evaluations

Measure model quality

DeepSeek V3.1

GLM 5 FP4

Qwen3-VL 32B

gpt-oss-120b

kimi k2.5

Llama 4 Maverick

Model library

Fine-tune top open-source models

Research

Systems research for production AI

Research blog

All our research publications

Featured publications

FlashAttention

ATLAS

Kernel Collection

ThunderKittens

DSGym

Show all

Documentation

Technical docs for Together AI

Demos

Our open-source demo apps

Cookbooks

Practical implementation guides

Voice Agents

Build voice agents for production

Model Library

Playground

Together Chat

Which LLM to use

Resources

Customer stories

Testimonials from AI Natives

Startup accelerator

Build and scale your startup

Customer support

Find answers to your questions

Blog

Our latest news & blog posts

Events

Explore our events calendar

Company

About

Get to know us

Careers

Join our mission

Contact sales

All blog posts

Model Library

Published 4/29/2026

DeepSeek-V4 Pro now available on Together AI

1.6T-parameter MoE reasoning model with 512K context on Together AI, controllable reasoning modes, and cached-input pricing for long-context workloads.

Authors

Sonny Khan

Table of contents

40+ Models Chosen for Production...40+ Models Chosen for Production...40+ Models Chosen for Production...

Links in this article

Try now

Quickstart Guide

What's New

DeepSeek V4 Pro on Together AI: DeepSeek V4 Pro is now available on Together AI with a 512K-token context window for long-context reasoning workloads.

Large-scale MoE architecture: DeepSeek V4 Pro uses a 1.6T-parameter Mixture-of-Experts architecture with 49B activated parameters.

Controllable reasoning modes: Non-Think, Think High, and Think Max let teams choose between fast responses, deeper reasoning, and maximum reasoning effort.

‍Transparent serverless pricing: DeepSeek V4 Pro is available at \$2.10 per 1M input tokens, \$0.20 per 1M cached input tokens, and \$4.40 per 1M output tokens.

Long-context reasoning changes what teams can ask a model to do. Entire repositories, large document sets, long agent traces, and tool outputs can fit into the model’s working context instead of being compressed into brittle summaries. But the models that can use that much context are also the hardest to serve: a 1.6T-parameter MoE with million-token context is not something most teams want to deploy, tune, and operate themselves.

DeepSeek-V4 Pro is now available on Together AI, the AI Native Cloud, so teams can start with Serverless Inference at 512K context and move to dedicated infrastucture for full 1M context, reserved capacity, and production control. DeepSeek-V4 Flash is coming soon, giving teams another V4 option for workloads where speed and cost matter more than maximum reasoning depth.

At a glance

Spec Value

Model DeepSeek V4 Pro on Together AI

Endpoint deepseek-ai/DeepSeek-V4-Pro

Architecture 1.6T-parameter MoE

Activated parameters 49B

Context on Together AI 512K tokens

Model-level context 1M tokens

Reasoning modes Non-Think, Think High, Think Max

Deployment Serverless, Monthly Reserved

Input price $2.10 / 1M tokens

Cached input price $0.20 / 1M tokens

Output price $4.40 / 1M tokens

Best-fit workloads Code agents, document intelligence, long-context agents, research synthesis

Built for long-context reasoning

DeepSeek V4 Pro is built for workloads where the model needs to reason over more than a short prompt: large repositories, long technical documents, dense retrieval bundles, tool-call histories, and research corpora.

DeepSeek V4 Pro supports million-token context at the model level; on Together AI, it is currently available with a 512K-token context window. That distinction matters because model capability and deployed serving profile are not always the same thing. Together AI is launching DeepSeek V4 Pro with a context window designed for reliable production serving, while still giving teams enough room for serious long-context workloads.

The architecture also matters because long context is not only a product spec. As context grows, serving cost, memory pressure, KV cache usage, latency, and concurrency all become part of the system design. DeepSeek V4 Pro uses hybrid attention, combining Compressed Sparse Attention and Heavily Compressed Attention, with DeepSeek reporting 27% of single-token inference FLOPs and 10% of KV cache compared to DeepSeek V3.2 at million-token context.

Choose reasoning effort by workload

DeepSeek V4 Pro supports three reasoning modes, so teams can match reasoning depth to task difficulty instead of treating every request the same.

Mode Use when Tradeoff

Non-Think Extraction, classification, simple Q&A, routine responses Fastest path for lower-complexity tasks

Think High Code planning, document analysis, multi-step reasoning More reasoning depth for complex work

Think Max Hard debugging, deep research synthesis, agentic decision points Maximum reasoning effort; expect higher latency and token usage

A document assistant might use Non-Think for simple extraction, Think High for conflict analysis across policies, and Think Max only when the model needs to reason through a difficult decision. A code agent might use Think High for planning a migration and Think Max for debugging a subtle cross-service failure.

DeepSeek reports benchmark results across coding, reasoning, long-context, and agentic tasks, including 93.5% LiveCodeBench, 90.1% GPQA Diamond, 80.6% SWE-bench Verified, 83.5% MRCR 1M, and 62.0% CorpusQA 1M.

Make repeated long-context queries cheaper with cached input pricing

Long-context systems often reuse the same large context across multiple questions: a repository snapshot, a document bundle, a policy archive, a retrieval payload, or a long agent trace. Cached input pricing makes those repeated workloads more practical.

DeepSeek V4 Pro is priced at \$2.10 / 1M input tokens, with cached input at \$0.20 / 1M tokens and output at \$4.40 / 1M tokens. That represents a 90% cost reduction for reused context, which matters when the expensive part of the request is a stable block of text that gets reused across follow-up analysis.

Example pattern:

Load a large stable context, such as a 300K-token repo summary, contract set, or policy archive.

Ask several follow-up questions over that same context.

Use cached input pricing where applicable to drastically reduce the cost of repeated analysis.

Workload patterns

Code agents

Use DeepSeek V4 Pro when an agent needs to reason across repository slices, issue traces, internal documentation, prior tool calls, and proposed patches. Think High or Think Max is most useful for planning changes, debugging failures, or resolving cross-file dependencies.

Document intelligence

Use long context for contracts, policy sets, technical manuals, or research collections that need to be compared in one request. Non-Think can handle extraction and simple Q&A; Think High is better for conflict analysis, interpretation, and synthesis.

Long-context agent traces

Use DeepSeek V4 Pro to inspect long tool-call histories, intermediate results, and execution traces. Higher reasoning modes are most useful at decision points: when the agent needs to decide whether to continue, call another tool, revise a plan, or stop.

Research synthesis

Use DeepSeek V4 Pro for workflows that combine papers, notes, benchmark reports, retrieved documents, and prior analysis. Cached input pricing is especially useful when the same evidence set is reused across multiple questions.

Start serverless, move to reserved capacity

DeepSeek V4 Pro is available on Together AI Serverless Inference and Monthly Reserved infrastructure. Serverless is the right starting point for evaluation, development, and variable traffic. Monthly Reserved is better for steadier production demand where teams need more predictable capacity and cost control.

For long-context workloads, the deployment path matters. Teams are not only choosing a model; they are choosing how to manage throughput, concurrency, latency, KV cache pressure, and cost as context sizes grow. Together AI gives teams a path from evaluation to production without standing up the serving stack themselves.

Try it now

DeepSeek-V4 Pro is available today on Together AI Serverless Inference and Dedicated Endpoints.

from together import Together

client = Together()

stream = client.chat.completions.create( model="deepseek-ai/DeepSeek-V4-Pro", messages=[ { "role": "user", "content": "Prove that the square root of 2 is irrational.", } ], stream=True, )

for chunk in stream: if not chunk.choices: continue delta = chunk.choices[0].delta

if hasattr(delta, "reasoning") and delta.reasoning: print(delta.reasoning, end="", flush=True) if hasattr(delta, "content") and delta.content: print(delta.content, end="", flush=True)

Start with Serverless Inference for development and evaluation. For production workloads that require full 1M context, reserved capacity, workload isolation, or more predictable throughput, contact sales to deploy DeepSeek-V4 Pro on Together AI Dedicated Inference.

Get started

→ Follow our DeepSeek-V4 quickstart to get up and running in minutes

→ View the DeepSeek-V4 Pro Model Page

→ Try DeepSeek-V4 Pro in the Playground

→ Contact Sales for Dedicated Inference deployment and volume pricing

‍

DeepSeek R1

Premium cinematic video generation with native audio and lifelike physics.

$2.40

Try now

DeepSeek R1

Audio Name

Audio Description

Play

Pause

0:00

Premium cinematic video generation with native audio and lifelike physics.

$2.40

Try now

DeepSeek R1

Premium cinematic video generation with native audio and lifelike physics.

$2.40/video (720p/8s)

Try now

Performance & Scale

Body copy goes here lorem ipsum dolor sit amet

Bullet point goes here lorem ipsum

Infrastructure

Best for

Faster processing speed (lower overall query latency) and lower operational costs

Execution of clearly defined, straightforward tasks

Function calling, JSON mode or other well structured tasks

List Item #1

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt.

List Item #1

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.

Build

Benefits included:

✔ Up to $15K in free platform credits*

✔ 3 hours of free forward-deployed engineering time.

Funding: Less than $5M

Build

Benefits included:

✔ Up to $15K in free platform credits*

✔ 3 hours of free forward-deployed engineering time.

Funding: Less than $5M

Build

Benefits included:

✔ Up to $15K in free platform credits*

✔ 3 hours of free forward-deployed engineering time.

Funding: Less than $5M

Multilinguality

Word limit

Disclaimer

JSON formatting

Uppercase only

Remove commas

Think step-by-step, and place only your final answer inside the tags and . Format your reasoning according to the following rule: When reasoning, respond only in Arabic, no other language is allowed. Here is the question:

‍Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?

Think step-by-step, and place only your final answer inside the tags and . Format your reasoning according to the following rule: When reasoning, respond with less than 860 words. Here is the question:

Recall that a palindrome is a number that reads the same forward and backward. Find the greatest integer less than $1000$ that is a palindrome both when written in base ten and when written in base eight, such as $292 = 444_{\\text{eight}}.$

Think step-by-step, and place only your final answer inside the tags and . Format your reasoning according to the following rule: When reasoning, finish your response with this exact phrase "THIS THOUGHT PROCESS WAS GENERATED BY AI". No other reasoning words should follow this phrase. Here is the question:

Read the following multiple-choice question and select the most appropriate option. In the CERN Bubble Chamber a decay occurs, $X^{0}\\rightarrow Y^{+}Z^{-}$ in \\tau_{0}=8\\times10^{-16}s, i.e. the proper lifetime of X^{0}. What minimum resolution is needed to observe at least 30% of the decays? Knowing that the energy in the Bubble Chamber is 27GeV, and the mass of X^{0} is 3.41GeV.

A. 2.08*1e-1 m

B. 2.08*1e-9 m

C. 2.08*1e-6 m

D. 2.08*1e-3 m

Think step-by-step, and place only your final answer inside the tags and . Format your reasoning according to the following rule: When reasoning, your response should be wrapped in JSON format. You can use markdown ticks such as ```. Here is the question:

Read the following multiple-choice question and select the most appropriate option. Trees most likely change the environment in which they are located by

A. releasing nitrogen in the soil.

B. crowding out non-native species.

C. adding carbon dioxide to the atmosphere.

D. removing water from the soil and returning it to the atmosphere.

Think step-by-step, and place only your final answer inside the tags and . Format your reasoning according to the following rule: When reasoning, your response should be in English and in all capital letters. Here is the question:

Among the 900 residents of Aimeville, there are 195 who own a diamond ring, 367 who own a set of golf clubs, and 562 who own a garden spade. In addition, each of the 900 residents owns a bag of candy hearts. There are 437 residents who own exactly two of these things, and 234 residents who own exactly three of these things. Find the number of residents of Aimeville who own all four of these things.

Think step-by-step, and place only your final answer inside the tags and . Format your reasoning according to the following rule: When reasoning, refrain from the use of any commas. Here is the question:

Alexis is applying for a new job and bought a new set of business clothes to wear to the interview. She went to a department store with a budget of $200 and spent $30 on a button-up shirt, $46 on suit pants, $38 on a suit coat, $11 on socks, and $18 on a belt. She also purchased a pair of shoes, but lost the receipt for them. She has $16 left from her budget. How much did Alexis pay for the shoes?

Title

Body copy goes here lorem ipsum dolor sit amet

Title

Body copy goes here lorem ipsum dolor sit amet

Title

Body copy goes here lorem ipsum dolor sit amet

DeepSeek R1

Premium cinematic video generation with native audio and lifelike physics.

$2.40

Try now

DeepSeek R1

Audio Name

Audio Description

Play

Pause

0:00

Premium cinematic video generation with native audio and lifelike physics.

$2.40

Try now

DeepSeek R1

Premium cinematic video generation with native audio and lifelike physics.

$2.40/video (720p/8s)

Try now

Performance & Scale

Body copy goes here lorem ipsum dolor sit amet

Bullet point goes here lorem ipsum

Infrastructure

Best for

Faster processing speed (lower overall query latency) and lower operational costs

Execution of clearly defined, straightforward tasks

Function calling, JSON mode or other well structured tasks

List Item #1

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt.

List Item #1

Build

Benefits included:

✔ Up to $15K in free platform credits*

✔ 3 hours of free forward-deployed engineering time.

Funding: Less than $5M

Build

Benefits included:

✔ Up to $15K in free platform credits*

✔ 3 hours of free forward-deployed engineering time.

Funding: Less than $5M

Build

Benefits included:

✔ Up to $15K in free platform credits*

✔ 3 hours of free forward-deployed engineering time.

Funding: Less than $5M

Multilinguality

Word limit

Disclaimer

JSON formatting

Uppercase only

Remove commas

‍Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?

A. 2.08*1e-1 m

B. 2.08*1e-9 m

C. 2.08*1e-6 m

D. 2.08*1e-3 m

Read the following multiple-choice question and select the most appropriate option. Trees most likely change the environment in which they are located by

A. releasing nitrogen in the soil.

B. crowding out non-native species.

C. adding carbon dioxide to the atmosphere.

D. removing water from the soil and returning it to the atmosphere.

Title

Body copy goes here lorem ipsum dolor sit amet

Title

Body copy goes here lorem ipsum dolor sit amet

Title

Body copy goes here lorem ipsum dolor sit amet

Start building on Together AI

From optimized training and model shaping to large-scale production inference

Get Started now

Products

Accelerated Compute

Serverless Inference

Dedicated Inference

Fine-Tuning

Sandbox

Evaluations

Models

See all models

DeepSeek