Together AI

2,648 posts

Together AI

@togethercompute

Accelerate inference, model shaping, and pre-training on a research-optimized platform.

San Francisco, CAtogether.ai

Joined November 2022

396 Following

54.9K Followers

Together AI’s posts

Pinned

Together AI

@togethercompute

Apr 25

Introducing DeepSeek V4 Pro, a long-context model with hybrid attention, three reasoning modes, and SOTA coding performance. AI natives can now use DeepSeek V4 Pro on Together AI and benefit from reliable inference for long-horizon coding and agentic workflows.

spent 20 years studying how language forms intelligence. When LLMs exploded, he saw something everyone missed. These systems were fluent, confident, and wrong. And no one could tell the difference.

Switching to Together AI flipped it:

Zero training-blocking failures

~50% cost savings vs. AWS

Issues resolved within hours via shared Slack They could finally focus on building the model.

The AI industry is racing toward superintelligence. Dan and Scaled Cognition built something more : Super-Reliable Intelligence. "The smartest model in the world is useless if it can’t get the answer right every single time." Read the full story here:

How Scaled Cognition Trains APT-1 on Together AI GPU Clusters

On 5/5

and team will discuss DSV4’s hybrid attention and KV cache efficiency, should be a great session!

Quote

Together AI

@togethercompute

May 1

Join us Tue 5/5: #DeepSeek-V4's hybrid attention + sparse MoE reduces KV cache up to 90%, enabling 1M-token context. We'll cover why that makes it great for agentic workflows, what it took to serve at scale, and how to build with it. Hear from @realDanFu @JueWANG26088228

Most model trainings outside of frontier labs fail.

Because of bad or insufficient data.

Or just data for what you want + not general capabilities.

Most builders give up + become elevated prompt engineers. Today,

@adaption_ai

@togethercompute

start fixing that.

Quote

adaption

@adaption_ai

Apr 30

We believe that intelligence should not arrive preconfigured. @togethercompute is now available directly inside the Adaption platform, connecting Adaptive Data with large-scale training in a single workflow. One platform, end to end. Stop inheriting intelligence. Shape it.

0:04 / 0:29

We’re excited to partner with

@adaption_ai

to make Together Fine-Tuning natively available in Adaptive Data. AI natives can now move from optimized training data in Adaptive Data to fine-tuned open models on Together AI.

adaption and Sara Hooker

Fine-tuning quality starts before the training run. Adaptive Data helps teams analyze, adapt, and improve datasets; Together Fine-Tuning turns those shaped datasets into specialized open models.

With this integration, teams can optimize data, launch fine-tuning, evaluate results, and deploy on Together AI inference through a tighter workflow. Learn more:

Announcing Together AI and Adaption Partnership

We believe that intelligence should not arrive preconfigured.

@togethercompute

is now available directly inside the Adaption platform, connecting Adaptive Data with large-scale training in a single workflow. One platform, end to end. Stop inheriting intelligence. Shape it.

$3/million output tokens. Qwen 3.5 Plus is basically a frontier model. Let that sink in.

Quote

Together AI

@togethercompute

Apr 30

Introducing Qwen3.6-Plus from @Alibaba_Qwen, a 1M-context model built for real-world agents, agentic coding, and multimodal reasoning. AI natives can now use Qwen3.6-Plus on Together AI and benefit from reliable inference for production-scale agent workflows.

Highlights:

1M context for long-horizon agentic workflows

Stronger agentic coding across frontend, repo-level, and terminal-based tasks

Multimodal reasoning across text, image, and video inputs

Production-ready on the AI Native Cloud—serverless inference at $0.50

Qwen3.6-Plus is now available on Together AI Try it now:

together.ai

Qwen3.6-Plus API | Together AI

Multimodal model that turns UI screenshots and design mockups into production code, with 1M context, thinking mode, and strong agentic coding performance.

Introducing Qwen3.6-Plus from

@Alibaba_Qwen

, a 1M-context model built for real-world agents, agentic coding, and multimodal reasoning. AI natives can now use Qwen3.6-Plus on Together AI and benefit from reliable inference for production-scale agent workflows.

Qwen and Alibaba Cloud

Last week we announced DeepSeek-V4. Today we’re sharing a closer look at DeepSeek-V4 Pro on Together AI: 512K context, controllable reasoning modes, and cached-input pricing for long-context workloads. Read more:

DeepSeek-V4 Pro now available on Together AI

From together.ai