Together AI
Together AI is an AI cloud platform for running, fine-tuning, and training open-source foundation models at high performance and low cost. Founded in June 2022 by Vipul Ved Prakash, Ce Zhang, Chris Re, and Percy Liang, it serves over 200 open-source models through an OpenAI-compatible API to more than one million developers (as of March 2026) and operates its own vertically integrated GPU cloud [2][24]. It reported 200 MW of secured power capacity in February 2025; by mid-2026 it cited commitments for more than 500 MW of compute capacity, plus a planned 250 MW data center in Saudi Arabia with HUMAIN [5]. By early 2026 the company had reportedly reached roughly $1 billion in annualized revenue [12]. On July 1, 2026 it announced an $800 million Series C led by Aramco Ventures at an $8.3 billion post-money valuation, making it one of the largest independent infrastructure providers built around open-weight AI models [21][22].
The company sits at the center of a fast-growing market for managed access to open-weight models like Llama, DeepSeek, Qwen, and Mistral, competing with specialized inference providers such as Fireworks AI and Replicate, and with hyperscaler offerings like Amazon Bedrock and Google Vertex AI. Its strategy combines research-grade optimization, including kernels developed by chief scientist Tri Dao, with a vertically integrated GPU cloud designed to capture both inference traffic and training workloads from AI-native startups and large enterprises. "Our AI Acceleration Cloud uniquely provides organizations with the performance, security, and functionality required to train frontier models and build production-scale AI applications with incredible cost efficiency," CEO Vipul Ved Prakash said when announcing the company's Series B [17].
What is Together AI?
Together AI is a full-stack "AI Acceleration Cloud" that lets developers and enterprises access open-source models through a single API and reserve dedicated GPU capacity for training and inference. Where most rivals do either serverless model serving or raw GPU rental, Together AI does both on infrastructure it increasingly owns and operates, and pairs it with research-derived performance optimizations. The platform spans serverless inference, fine-tuning, custom training, GPU clusters, code-execution sandboxes, evaluations, and data preparation, all built around the open-weight model ecosystem rather than a single proprietary model.
History and founding
Together AI (originally incorporated as Together Computer Inc.) was founded in June 2022 to address what its founders saw as a growing compute moat limiting access to the hardware and infrastructure needed to train and deploy large language models. The founding team brought together industry experience and academic AI research [3].
Vipul Ved Prakash, who serves as CEO, previously co-founded and led Topsy (acquired by Apple) and Cloudmark (acquired by Proofpoint), both of which dealt with large-scale data processing. Ce Zhang, who serves as CTO, came from ETH Zurich, where his research focused on data management for machine learning and distributed training. Chris Re and Percy Liang are both professors at Stanford University, with Re's lab having produced foundational work on data-centric AI and Liang leading the Center for Research on Foundation Models (CRFM) [3][13].
The company initially explored a decentralized cloud model that would aggregate idle compute across data centers and coordinate high-latency networks for synchronized training. This research direction yielded several early publications on decentralized training over commodity internet links, but Together AI ultimately shifted toward a more conventional cloud infrastructure approach, building and operating its own high-bandwidth GPU clusters tuned for large model workloads.
A pivotal moment came in mid-2023, when Together AI brought on Tri Dao as chief scientist. Dao is the creator of FlashAttention and FlashAttention-2, memory-efficient attention algorithms that have become standard components in modern transformer training and inference stacks used by OpenAI, Anthropic, Meta, and Mistral. His research forms the basis for much of Together AI's performance advantage and underpins the Together Kernel Collection that ships with the platform's training and inference engines [3][13].
Founding team
| Founder | Role | Background |
|---|---|---|
| Vipul Ved Prakash | Co-founder, CEO | Founder of Topsy (acquired by Apple), Cloudmark (acquired by Proofpoint) |
| Ce Zhang | Co-founder, CTO | Former associate professor at ETH Zurich, distributed ML systems research |
| Chris Re | Co-founder, advisor | Professor at Stanford, founder of Snorkel AI, MacArthur Fellow |
| Percy Liang | Co-founder, advisor | Stanford professor, director of CRFM, co-author of HELM benchmark |
| Tri Dao | Chief scientist (joined 2023) | Creator of FlashAttention and Mamba, assistant professor at Princeton |
Together AI is headquartered in San Francisco. The full-time research team has grown to include other prominent open-source contributors and former hyperscaler infrastructure engineers, with a deliberate emphasis on hiring people who can move between systems research and product engineering.
RedPajama and open dataset work
Before Together AI became known primarily as an inference and training cloud, the company spent much of 2023 and 2024 publishing open datasets and pretraining recipes under the RedPajama project. RedPajama is a multi-institution effort involving Together, Ontocord.ai, ETH DS3Lab, Stanford CRFM, and Hazy Research that set out to make state-of-the-art language model pretraining fully reproducible without reliance on closed corpora [11].
The first release, RedPajama-V1, reconstructed the training mixture described in the original LLaMA paper and contained roughly 1.2 trillion tokens spanning Common Crawl, GitHub, books, ArXiv, Wikipedia, and Stack Exchange. The follow-on dataset, RedPajama-V2, is a web-only corpus containing more than 100 billion documents drawn from 84 Common Crawl snapshots, processed with the CCNet pipeline, and shipped together with 30 billion documents of quality signals and 20 billion deduplicated documents. Across both releases the datasets comprise more than 100 trillion tokens [11].
RedPajama datasets have been used directly or indirectly in the pretraining of several open models that later went into production, including Snowflake Arctic, Salesforce XGen, and AI2 OLMo. The associated tooling lives on GitHub under the togethercomputer organization, and the data itself is mirrored on Hugging Face under a permissive license. The project doubled as a credibility builder for Together AI's training services, since enterprise customers could see the company contributing to and curating the same kinds of corpora used for foundation model training.
In 2024, Together AI published the RedPajama-V2 paper at NeurIPS, formalizing the data preparation pipeline and analyzing the impact of different quality signals on downstream model quality. The work has since been cited in research from major labs and has informed how Together AI structures its training data services for paying customers.
How much funding has Together AI raised?
Together AI has raised significant venture capital across multiple rounds, reflecting investor confidence in the infrastructure layer of the AI ecosystem. Its Series A, Series B and July 2026 Series C alone total more than $1.2 billion, and its valuation rose from roughly $500 million in late 2023 to $8.3 billion in July 2026 [4][2][21].
| Round | Date | Amount | Lead investors | Valuation |
|---|---|---|---|---|
| Seed | May 2023 | $20M | Lux Capital [27] | N/A |
| Series A | November 2023 | $102.5M | Kleiner Perkins | ~$500M |
| Series A extension | March 2024 | $106M | Salesforce Ventures | $1.25B |
| Series B | February 2025 | $305M | General Catalyst, Prosperity7 | $3.3B |
| Series C | July 2026 | $800M | Aramco Ventures | $8.3B (post-money) |
The Series A round in November 2023 included participation from NVIDIA, NEA, Prosperity7 Ventures, Greycroft, and 137 Ventures, among others [4]. The March 2024 Series A extension led by Salesforce Ventures pushed Together AI's valuation past the unicorn threshold to $1.25 billion and brought Salesforce into the cap table as both an investor and a customer.
The $305 million Series B announced on February 20, 2025 was led by General Catalyst and co-led by Prosperity7, the venture arm of Saudi Aramco. The round valued Together AI at $3.3 billion and brought in additional capital from Salesforce Ventures, DAMAC Capital, Nvidia, Kleiner Perkins, March Capital, Emergence Capital, Lux Capital, SE Ventures, Greycroft, Coatue, Definition, Cadenza Ventures, Long Journey Ventures, Brave Capital, Scott Banister, SK Telecom, and former Cisco CEO John Chambers. Proceeds were earmarked for expanding GPU cluster capacity and deploying NVIDIA Blackwell GPUs across multiple data centers in North America [2][17]. General Catalyst managing director Marc Bhargava said at the time that "Vipul and team have built an incredible tech platform and business, emerging as a dominant player in AI infrastructure in less than two years" [17].
In March 2026, multiple outlets reported that Together AI was in talks to raise approximately $1 billion at a $7.5 billion pre-money valuation, more than doubling its Series B mark just thirteen months earlier. Prosperity7 was again reported as a potential lead, with significant follow-on participation expected from existing strategic investors including Nvidia. The talks were tied to the company's roughly $1 billion annualized revenue run rate and its expanding Blackwell deployment [12]. The round closed on different terms. On July 1, 2026 Together AI announced an $800 million Series C at an $8.3 billion post-money valuation, led by Aramco Ventures with participation from Vista Equity Partners, General Catalyst, Emergence Capital, NVIDIA, March Capital, Pegatron and S Ventures (SentinelOne), among others [21][22]. TechCrunch reported that the company claimed annual bookings of more than $1.15 billion as of its most recent quarter [21].
Revenue and growth
Together AI has scaled revenue at a pace reminiscent of hyperscaler AI services. Annualized revenue rose from approximately $130 million at the end of 2024 to $300 million in September 2025, and reached approximately $1 billion by February 2026, more than three times its mid-2025 level [5][12].
| Period | Annualized revenue | Notes |
|---|---|---|
| End of 2024 | ~$130M | Per Sacra estimates |
| September 2025 | ~$300M | Per Sacra and LinkedIn reporting |
| Early 2026 | ~$1B | Reported by The Information ahead of $7.5B valuation talks |
At the Series B, the company said it had grown its user base to over 450,000 AI developers during 2024 [17]. Revenue is generated through two primary lines. The first is per-token API usage on the serverless inference platform, which accounts for roughly 30 to 40 percent of revenue. The second and larger share comes from renting GPU capacity, whether as dedicated inference endpoints, reserved GPU clusters, or on-demand training infrastructure. As Blackwell capacity has come online, the GPU rental business has scaled rapidly, with several customers committing to multi-month reservations for thousands of GPUs at a time [12].
Key products and services
Together AI organizes its offerings around four pillars: serverless inference, fine-tuning, custom training, and GPU clusters, with additional managed services for code execution, evaluations, and data preparation that wrap the core compute layer.
Inference API
Together AI's inference API provides OpenAI-compatible access to over 200 open-source models. The platform is built on a proprietary inference engine that incorporates FlashAttention-3 kernels and advanced quantization techniques, delivering what the company claims is 2 to 3 times faster inference than hyperscaler solutions [2][17].
The API supports chat completions, text completions, embeddings, image generation, and audio processing. Its OpenAI API compatible format means that developers can often switch from OpenAI or other providers with minimal code changes, simply swapping the base URL and API key.
How fast is Together AI inference?
Together AI's focus on inference speed has resulted in measurable performance advantages over competing platforms. The company reports up to 2x faster inference for top open-source models like Qwen, DeepSeek, and Kimi through a combination of GPU-level optimization, advanced speculative decoding, and FP4 quantization [9].
Key performance technologies include:
| Technology | Description | Impact |
|---|---|---|
| FlashAttention-3 kernels | Memory-efficient attention optimized for latest GPU architectures | Reduced memory overhead, higher throughput |
| FP4 quantization | 4-bit floating-point model compression | Lower memory usage, faster inference with minimal quality loss |
| ATLAS (AdapTive-LeArning Speculative System) | Learns from production traffic to predict and pre-generate tokens | Acceleration beyond static optimization |
| Together Kernel Collection (TKC) | Proprietary CUDA kernels optimized for open-source model architectures | Model-specific performance gains |
| Speculative decoding | Small draft model predicts tokens, verified by the larger model | 1.5-2x throughput improvement for autoregressive generation |
On NVIDIA Blackwell architecture, Together AI ranks first in independent speed benchmarks for several top open-source models [9]. The ATLAS system is particularly notable because it continuously adapts to real-world production traffic rather than relying solely on static optimizations, meaning performance improves over time as the system observes usage patterns. In July 2024 Together AI said its Inference Engine 2.0 delivered decoding throughput up to four times faster than open-source vLLM on Meta Llama 3 models [26].
Supported models
Together AI maintains one of the broadest catalogs of open-source models available through a managed inference service. New models are typically added within days of public release, and the platform aims to support every major open-weight model family at parity with the original authors' reference implementations.
| Model family | Provider | Notable variants | Use cases |
|---|---|---|---|
| Llama 3.1 / 3.3 / Llama 4 | Meta | Llama 3.1 8B and 405B, Llama 3.3 70B, Llama 4 Maverick | General text, code, multilingual |
| Mistral / Mixtral | Mistral AI | Mistral-7B, Mixtral-8x22B, Mistral Small 3, Mistral Large | Text, code, RAG |
| Qwen 2.5 / Qwen3 | Alibaba | Qwen 2.5, Qwen3, Qwen3-Coder-Next, Qwen3.5-397B | General text, code, multilingual |
| DeepSeek | DeepSeek | DeepSeek R1, DeepSeek-V3, DeepSeek-V3.1 | Reasoning, code, research |
| Gemma | Gemma 2 9B, Gemma 2 27B, Gemma 3n E4B | Lightweight text generation | |
| DBRX | Databricks | DBRX Instruct | Enterprise text generation |
| Stable Diffusion | Stability AI | SDXL, Stable Diffusion 3 | Image generation |
| Whisper | OpenAI | Whisper Large v3 | Speech-to-text |
| Refuel LLM-2 | Refuel.ai (Together AI) | Refuel-LLM-2 | Data labeling and structured extraction |
The platform also supports specialized models for code generation, vision-language tasks, and embeddings. Following the May 2025 acquisition of Refuel.ai (covered below), Refuel LLM-2 is offered as a first-party data-task model with both serverless inference and LoRA fine-tuning support [10].
Fine-tuning
Together AI's fine-tuning service allows users to customize open-source models on their own data. The managed service supports supervised fine-tuning (SFT) and direct preference optimization (DPO). In March 2026, Together AI expanded fine-tuning with native support for tool call training, reasoning model fine-tuning, and vision-language model adaptation, along with more efficient support for models above 100 billion parameters [28].
The fine-tuning pipeline includes cost and ETA estimates before job submission, and Together AI reports up to 6 times higher throughput compared to earlier versions of its training infrastructure [28]. Customers can run jobs either inside the managed fine-tuning service or by reserving raw GPU capacity and using the open-source Together training scripts directly.
In October 2026 Together AI launched Together Post-Training, a lower-level alternative to the managed fine-tuning service. The customer defines the whole training loop for reinforcement learning or supervised fine-tuning, while Together runs the trainer and rollout generators on its GPUs and bills each model resource per GPU-minute. Scripts written for the Tinker API from Thinking Machines Lab can be moved over by changing one import line [23].
Fine-tuning pricing and methods
Together AI prices fine-tuning per million training tokens, with a separate rate for each base model and a minimum charge per job. Rates listed in September 2026 included [19]:
| Base model (examples) | LoRA SFT | LoRA DPO | Full SFT | Full DPO |
|---|---|---|---|---|
| Llama 3.1 8B, Qwen3.5 up to 9B | $0.34 | $0.84 | $0.38 | $0.94 |
| Qwen3.5 27B, Gemma 4 31B, Mixtral 8x7B | $1.05 | $2.62 | $1.16 | $2.88 |
| Llama 3.3 70B | $2.03 | $5.08 | $2.24 | $5.60 |
| Llama 4 Maverick | $5.60 | $14.00 | Not listed | Not listed |
| DeepSeek-V3.1 | $7.00 | $17.50 | Not listed | Not listed |
Minimum charges ranged from $4 to $60 per job depending on the model [19].
LoRA (Low-Rank Adaptation) fine-tuning is more cost-effective than full fine-tuning because it updates only a small subset of model parameters. For most use cases, LoRA provides comparable quality at lower cost, while full fine-tuning offers maximum customization for highly specialized applications.
Custom training
For organizations that need to train models from scratch or do extensive continued pretraining, Together AI offers custom training infrastructure built on its GPU clusters. This service targets AI-native companies and research labs that need direct access to large-scale compute and bespoke parallelism strategies.
Training performance has improved significantly with the introduction of Blackwell GPUs. Training a 70B-parameter Llama-architecture model in BF16 precision with an optimized TorchTitan plus Together Kernel Collection stack reached 15,264 tokens per second per NVIDIA HGX B200 node (8 GPUs), up from 8,080 tokens per second per HGX H100 node, which Together described as a 90 percent improvement in training throughput [25]. Together AI's training services are designed to compose with PyTorch FSDP, Megatron-LM, and TorchTitan, with the kernel collection providing drop-in acceleration for the most expensive operators.
Together GPU Clusters
Together AI operates its own GPU cluster infrastructure, offering on-demand access to NVIDIA H100, H200, and (beginning in 2025) Blackwell B200 and GB200 GPUs. The company has secured 200 MW of power capacity across multiple data centers in North America, with a facility in Maryland that went live in July 2025 and planned capacity in Memphis, according to Sacra [5].
In September 2025, Together AI launched a self-service GPU infrastructure tier that allows customers to provision clusters ranging from a single eight-GPU node to multi-node systems with hundreds of processors [8]. The self-service model supports the latest NVIDIA Hopper and Blackwell hardware and is optimized for distributed training and elastic inference workloads.
Hypertec Cloud and the 36,000 GPU GB200 cluster
In November 2024, Together AI announced a partnership with Hypertec Cloud to co-build one of the world's largest optimized GPU clusters, featuring 36,000 NVIDIA GB200 NVL72 GPUs. Deployment began in Q1 2025 and is being layered on top of the thousands of H100 and H200 GPUs already operational across North America. Combined with previously announced commitments, the partnership gives Together AI secured data center capacity for over 100,000 GPUs throughout 2025 [14].
The GB200 NVL72 platform pairs Grace CPUs with Blackwell GPUs using fifth-generation NVLink, and Nvidia rates it as delivering up to 30x faster real-time inference for trillion-parameter models and up to 4x accelerated training versus the previous Hopper generation. The Hypertec deployment uses liquid cooling throughout and is integrated with Together AI's networking and orchestration stack, including the Together Kernel Collection [14].
GPU cluster pricing
GPU clusters are available across multiple commitment levels [7]. The rates below are per GPU-hour as listed on Together's pricing page in September 2026 [19]:
| GPU type | On-demand (per hour) | Dedicated inference (per hour) | Notes |
|---|---|---|---|
| NVIDIA HGX H100 (80GB) | $3.99 | $5.49 (promotional $3.99 through September 30, 2026) | Most widely available |
| NVIDIA HGX H200 (141GB) | $5.99 | No listed rate (contact sales) | Higher memory for larger models |
| NVIDIA HGX B200 (180GB) | $8.19 | $8.99 | Blackwell architecture |
| NVIDIA HGX B300 | $9.99 | No listed rate (contact sales) | Blackwell Ultra |
| NVIDIA GB200 NVL72 | Custom | Custom | Liquid-cooled rack-scale via Hypertec |
Reserved capacity with multi-month commitments offers reduced rates: in September 2026, 91-to-180-day reservations were listed at $3.19 (H100), $3.99 (H200), and $6.79 (B200) per GPU-hour, and preemptible capacity at $1.99 (H100), $2.99 (H200), $4.09 (B200), and $4.99 (B300) [19]. Volume pricing is available for large-scale deployments through direct negotiation with Together AI's enterprise sales team.
Together Code Sandbox and Code Interpreter
In 2025, Together AI introduced two products targeted at AI agent developers: Together Code Sandbox (TCS) and Together Code Interpreter (TCI) [15]. Both run on the same underlying micro-VM infrastructure but expose different developer surfaces.
Together Code Sandbox provides customizable virtual machines that serve as full development environments for AI applications, including persistent state, interactive shells, and real-time previews. Together Code Interpreter is session-based and exposes a simpler API for one-shot or short-lived code execution, designed to be called by LLMs as a tool during agentic workflows.
| Capability | Together Code Sandbox | Together Code Interpreter |
|---|---|---|
| Use case | Full IDE-like development environments | Stateless tool-call code execution |
| Persistence | Git-versioned filesystem | Session-scoped |
| VM sizing | Hot-swappable, 2 to 64 vCPUs, 1 to 128 GB RAM | Same micro-VM substrate |
| Cold start | 500 ms P95 from snapshot | Sub-second |
| Cloning | Under 1 second | Under 1 second |
| Typical caller | AI IDEs, SaaS platforms | Agent frameworks, RL training loops |
A notable production deployment is Agentica, an open-source project from Berkeley AI Research and the Sky Computing Lab, which used Together Code Interpreter during the training of DeepCoder-14B-Preview. Agentica reported running 1,024 code executions in parallel and scaling to more than 100 concurrent coding sandboxes with thousands of evaluations per minute, illustrating how TCI integrates with reinforcement learning operations for coding agents [15]. Together AI has positioned the sandbox products as direct alternatives to E2B and Modal for teams that want to keep both the model and the execution environment on the same platform.
Pricing for sandbox and other developer services
In addition to inference, fine-tuning, and GPU cluster pricing, Together AI publishes per-resource rates for ancillary services that are commonly bundled into agentic applications.
| Service | Pricing |
|---|---|
| Together Code Sandbox vCPU | $0.0446 per vCPU hour |
| Transcription (Whisper Large v3) | $0.0015 per audio minute |
| Storage | $0.16 per GiB per month |
| Evaluations dashboard | Included with platform |
Refuel.ai acquisition
On May 15, 2025, Together AI acquired Refuel.ai, a San Francisco startup founded in 2021 by Stanford alumni Rishabh Bhargava and Nihit Desai that used LLMs to clean, structure, and label enterprise data [10][18]. The acquisition price was not disclosed. Refuel.ai's flagship products included Refuel LLM-2, a family of models purpose-built for data labeling and structured extraction tasks that the company said achieved 50 percent fewer errors than state-of-the-art alternatives, and Refuel Cloud, a workflow platform for building multi-step data pipelines [18]. At the time of acquisition, Refuel was processing tens of millions of records and billions of tokens per week for customers including major financial institutions [10].
"Joining Together AI accelerates our mission to solve the data bottleneck that every AI team faces today," said Rishabh Bhargava, CEO of Refuel.ai [18]. The strategic rationale is straightforward: high-quality fine-tuning and post-training depend on high-quality data, and many Together AI customers were stitching together their own labeling pipelines on top of Together's inference and fine-tuning services. By bringing Refuel in-house, Together AI was able to offer end-to-end data preparation and customization inside a single platform.
What the acquisition added
| Capability | Pre-acquisition | Post-acquisition |
|---|---|---|
| First-party data labeling models | None | Refuel LLM-2 family with serverless inference and LoRA fine-tuning |
| Workflow orchestration | Notebook examples and custom scripts | Refuel Cloud workflows integrated with Together inference |
| Customer base | Developers and AI-native startups | Adds financial services and data-heavy enterprise accounts |
| Roadmap | Inference and training | Data preparation, labeling, evaluation, and agentic tooling |
The Refuel team joined Together AI's data and applied research groups, and Refuel LLM-2 was integrated into the Together model catalog so that existing customers could call it from the same OpenAI-compatible API used for Llama, DeepSeek, or Mistral models [10]. The acquisition was viewed by analysts as filling a notable gap relative to competitors like Databricks and Scale AI, both of which had moved aggressively to combine data infrastructure with model training services.
FlashAttention integration
Together AI's performance advantage is closely tied to its integration of FlashAttention, a family of memory-efficient attention algorithms created by chief scientist Tri Dao. FlashAttention reduces the memory overhead and computational cost of the attention mechanism in transformers, which is typically the primary bottleneck in both training and inference. FlashAttention is now used in production by OpenAI, Anthropic, Meta, and Mistral, but Together AI is the only commercial cloud whose product roadmap is shaped directly by its author [3][13].
The platform's inference engine incorporates FlashAttention-3 kernels, which are optimized for the latest NVIDIA GPU architectures including Hopper and Blackwell. Combined with advanced quantization (reducing model precision to lower memory usage and increase throughput) and the proprietary Together Kernel Collection, these optimizations enable the platform to serve large models at significantly lower latency and cost than standard serving frameworks [2]. Dao's work on Mamba and follow-on state space models has also informed Together AI's roadmap for serving non-transformer architectures efficiently.
Developer experience
Together AI has invested heavily in developer experience, making its platform accessible to developers already familiar with the OpenAI ecosystem.
Is Together AI compatible with the OpenAI API?
Yes. Together AI's API endpoints for chat, vision, images, embeddings, and speech are fully compatible with OpenAI's API format [11]. Developers using the OpenAI Python library, LangChain, or other frameworks with OpenAI integrations can point their existing applications at Together AI's servers by changing only the base URL and API key. All major parameters available in the OpenAI API work with Together AI, including streaming, function calling, and JSON mode.
This compatibility significantly reduces migration friction and allows developers to experiment with open-source models without rewriting application code. It also makes Together AI a common second provider in multi-provider routing setups, where applications fail over from a primary inference vendor to Together AI under load or for specific model families.
Developer tools and features
| Feature | Description |
|---|---|
| OpenAI-compatible API | Drop-in replacement for OpenAI endpoints |
| Code Sandbox | Isolated VM execution environment for code generation tasks |
| Code Interpreter | Session-based tool-call execution for agents |
| Evaluations dashboard | Compare model performance across benchmarks and custom datasets |
| Transcription API | Whisper-based speech-to-text |
| Streaming | Real-time token streaming for interactive applications |
| Function calling | Native tool use compatible with OpenAI function calling format |
| JSON mode | Guaranteed structured JSON output for programmatic consumption |
| Refuel data workflows | Multi-step labeling and extraction pipelines |
Pricing
Together AI's pricing is structured across its core product lines, with serverless inference priced per token and other services priced per GPU-hour or per resource-unit.
Serverless inference pricing
The rates below were listed earlier in 2026 unless marked otherwise. Together's serverless catalog changes often, and by November 2026 several of these models were no longer on its public serverless price list [7].
| Model | Input (per 1M tokens) | Output (per 1M tokens) | Notes |
|---|---|---|---|
| Llama 4 Maverick | $0.27 | $0.85 | Meta model released April 2025 |
| DeepSeek-V3.1 | $0.60 | $1.70 | Reasoning-focused |
| Qwen3.5-397B | $0.60 | $3.60 | Large multilingual model (November 2026 rate) |
| Mistral Small 3 | $0.10 | $0.30 | Efficient mid-range |
| Gemma 3n E4B | $0.02 | $0.04 | Ultra-lightweight |
| Llama 3.3 8B | $0.18 (combined) | - | Budget option |
| Qwen3-Coder-Next | $0.50 | $1.20 | Code-specialized |
| Refuel LLM-2 | Custom | Custom | Data labeling and extraction |
Other service pricing
| Service | Pricing |
|---|---|
| Fine-tuning (LoRA SFT, 8B-class models such as Llama 3.1 8B) | $0.34 per 1M tokens [19] |
| Fine-tuning (Full SFT, 8B-class models) | $0.38 per 1M tokens [19] |
| Dedicated inference (H100) | $5.49/hour ($3.99 promotional rate through September 30, 2026) [19] |
| Dedicated inference (B200) | $8.99/hour [19] |
| GPU clusters (H100, on-demand) | $3.99/hour [19] |
| GPU clusters (B200, on-demand) | $8.19/hour [19] |
| Together Code Sandbox | $0.0446 per vCPU hour |
| Storage | $0.16 per GiB per month |
The company positions its pricing as competitive with major cloud providers. Together AI offers a free tier that lets developers experiment without commitment, and enterprise customers can negotiate volume pricing.
How does Together AI compare with other inference providers?
Together AI competes against a mix of pure-play inference startups and hyperscaler offerings. The closest direct comparable is Fireworks AI, which also targets fast, low-cost serving of open-source models with its own proprietary inference engine (FireAttention). Replicate is broader in modality coverage but less optimized for production LLM throughput, while Hugging Face Inference Endpoints sit closer to the model hub itself. Hyperscaler offerings like Amazon Bedrock bundle access to a curated set of proprietary and open-source models inside the AWS account boundary [16].
| Feature | Together AI | Fireworks AI | Replicate | Hugging Face Inference | Amazon Bedrock |
|---|---|---|---|---|---|
| Focus | Open-source inference, training, GPU cloud | Open-source inference | Community model hosting | Model hub plus inference | Multi-provider managed AI |
| Custom training | Yes (GPU clusters) | Fine-tuning only | Fine-tuning only | AutoTrain for fine-tuning | Limited (fine-tuning only) |
| Inference engine | Together Kernel Collection, FlashAttention-3 | FireAttention | Standard serving | Standard serving | AWS-managed |
| Pricing model | Per-token plus per-GPU-hour | Per-token | Per-second compute | Per-token plus per-GPU-hour | Per-token |
| Code sandbox | Yes (TCS, TCI) | No | No | No | Yes (AgentCore Code Interpreter) [29] |
| Data labeling | Yes (Refuel.ai) | No | No | No | Limited |
| Target users | AI developers, enterprises | AI developers, startups | Indie developers, image/video apps | ML researchers, developers | Enterprise cloud users |
Together AI and Fireworks AI sit at the head of the speed-and-cost optimized inference category, with each company publishing benchmarks claiming a lead on different model families. Independent comparisons generally find that the two providers trade leadership depending on the model, batch size, and quantization regime, with Together AI more strongly differentiated on the training and GPU cluster side and on tooling like code sandboxes and Refuel data workflows. Replicate is most useful for image, video, and audio workloads where its long tail of community models and per-second billing fit a more experimental use case [16].
Enterprise customers and partnerships
Together AI has built a customer base that spans AI-native startups and large enterprises. Notable customers include Salesforce, Zoom, SK Telecom, Hedra, Cognition, Zomato, Krea, Cartesia, Pika Labs, and The Washington Post [2]. The company also counts Salesforce Ventures among its investors, reflecting a strategic partnership with one of the largest enterprise software companies.
The customer base reflects two distinct segments. AI-native startups like Cognition (maker of the Devin agent), Hedra (avatar and video generation), Krea (image and video generation), Pika Labs (video generation), and Cartesia (real-time voice with the Sonic model) use Together AI as their primary inference infrastructure, drawn by the performance advantages and cost savings that allow them to scale AI-intensive products without managing their own GPU clusters. Larger enterprises like Salesforce, Zoom, and SK Telecom typically use Together AI alongside their existing cloud infrastructure, often for specific workloads where open-source model performance and cost are critical factors.
Publicly reported customer outcomes include 24 percent faster training operations for Pika Labs and ultra-low latency voice AI through Cartesia's Sonic model integration. Krea has highlighted Together AI's ability to scale through traffic surges while maintaining performance during product launches.
Strategic partnerships
| Partner | Nature of relationship |
|---|---|
| Nvidia | Strategic investor, GPU supplier, joint go-to-market on Blackwell |
| Hypertec Cloud | Co-build partner on 36,000-GPU GB200 NVL72 cluster |
| Salesforce | Investor and enterprise customer |
| SK Telecom | Investor and Asia-Pacific customer |
| General Catalyst | Series B lead investor |
| Prosperity7 (Saudi Aramco) | Series B co-lead; Aramco Ventures led the July 2026 Series C [21] |
Infrastructure strategy
Together AI's approach to infrastructure represents a deliberate bet on vertical integration. Rather than renting capacity from major cloud providers, the company builds and operates its own GPU clusters, giving it direct control over hardware configuration, networking, and cooling. This approach mirrors the strategies of other AI-focused infrastructure companies like CoreWeave but is unusual among inference-focused platforms, which typically operate on top of existing cloud infrastructure [2].
The 200 MW of secured power capacity across multiple data centers is significant. For context, a single NVIDIA HGX B200 system consumes approximately 14 kilowatts, meaning 200 MW could theoretically power roughly 14,000 B200 systems at full load (before accounting for cooling and overhead). Combined with the Hypertec Cloud partnership and additional reserved capacity, Together AI said it had secured data center capacity for over 100,000 GPUs throughout 2025 [14].
The geographic distribution of data centers across North America, including facilities in Maryland and Memphis, provides redundancy and the ability to serve customers with data residency requirements.
Together AI also runs its inference stack on hyperscaler hardware where customers require it, including private deployments inside AWS, Microsoft Azure, and Google Cloud accounts. This hybrid posture lets the company chase wholesale GPU economics on its own metal while still meeting enterprise data residency and procurement constraints.
Open-source contributions
In parallel with its commercial products, Together AI has remained an active contributor to the open-source AI ecosystem. The company maintains or has contributed to several widely used projects:
| Project | Role | Description |
|---|---|---|
| RedPajama | Co-lead | 100 trillion token open dataset for LLM training |
| FlashAttention | Author (Tri Dao) | Memory-efficient attention used across the field |
| Mamba | Co-author (Tri Dao) | Selective state space architecture for sequence modeling |
| TorchTitan recipes | Contributor | Reference training recipes for large Llama-style models |
| Together Kernel Collection | Maintainer | CUDA kernels for transformer and SSM workloads |
| Refuel LLM-2 (post-acquisition) | Maintainer | Open-weight model for data labeling and extraction |
This posture matters commercially because most of Together AI's customers run open-weight models themselves, and a willingness to publish data, recipes, and kernels acts as both a recruiting tool and a credibility signal in the open-source community.
Current state (2025-2026)
As of May 2026, Together AI continues to scale rapidly. At its first AI Native Conf in March 2026, the company announced several product and business milestones [24]. The platform supports over 200 models across chat, image, audio, vision, code, and embeddings [17], and at the conference the company said it served thousands of customers and supported over one million developers [24]. Annualized revenue was reported at roughly $1 billion in early 2026 [12], and at its July 2026 Series C the company claimed annual bookings of more than $1.15 billion [21].
The competitive landscape for AI inference has intensified, with hyperscalers like AWS, Microsoft Azure, and Google Cloud all expanding their open-source model offerings, and specialized rivals like Fireworks AI, Anyscale, and Modal pushing on the same buyer base. Together AI's response has been to double down on performance and verticalization, investing in custom kernels, Blackwell GPU deployments, Refuel data workflows, code sandboxes, and vertically integrated infrastructure that gives it a cost advantage over larger but less specialized competitors.
The company's trajectory from a research-oriented startup to a multi-billion-dollar infrastructure provider illustrates the growing importance of the inference and training infrastructure layer in the AI ecosystem. With revenue growing rapidly, an expanding customer base, and a strong developer community, Together AI has established itself as one of the most prominent independent infrastructure companies serving the open-weight model market.
Continue reading
- ModelA model in machine learning is a representation used to map inputs to outputs, assign probabilities, describe relationships, or choose actions.Mentioned 83 times above
- OpenAIOpenAI is an American artificial intelligence research and deployment organization.Mentioned 17 times above
- Percy LiangPercy Liang is a professor of computer science at Stanford University, where he is the Thomas M. Siebel Professor of Machine LearningMentioned 3 times above
- Foundation ModelsA foundation model is a machine learning model trained on broad data, generally with self-supervision at scale, that can be adapted to a wide range of downstream tasks.Mentioned twice above
- Tri DaoTri Dao is a computer scientist who created the FlashAttention family of GPU attention algorithms and co-created the Mamba selective state-space architectureMentioned 6 times above
- Fireworks AIFireworks AI is an artificial intelligence infrastructure company that runs a high-performance inference platform for deploying and serving open large language models (LLMs), vision-language models, and…Mentioned 5 times above
More to read
References
- "Together AI." Together AI Official Website.
- ^1 ^2 ^3 ^4 ^5 ^6 ^7"Together AI Announces $305M Series B to Scale AI Acceleration Cloud." Together AI Blog, February 2025.
- ^1 ^2 ^3 ^4"Cloud Intelligence at the speed of 5000 tok/s - with Ce Zhang and Vipul Ved Prakash." Latent Space Podcast.
- ^1 ^2"Our $102.5M Series A." Together AI Blog, November 2023.
- ^1 ^2 ^3"Together AI revenue, valuation & funding." Sacra.
- "Build with leading AI models." Together AI Models Page.
- ^1 ^2"Pricing." Together AI.
- ^"Together AI launches self-service GPU infrastructure." SiliconANGLE, September 2025.
- ^1 ^2"Together AI delivers fastest inference for the top open-source models." Together AI Blog.
- ^1 ^2 ^3 ^4"Together AI acquires Refuel.ai to unlock data for developers and businesses building production-grade AI applications." Together AI Blog, May 2025.
- ^1 ^2 ^3"RedPajama: an Open Dataset for Training Large Language Models." arXiv:2411.12372.
- ^1 ^2 ^3 ^4 ^5"Nvidia Cloud Ally Together AI in Talks to Raise at $7.5 Billion Valuation." The Information, March 2026.
- ^1 ^2 ^3"About Us." Together AI Official Website.
- ^1 ^2 ^3"Together AI to Co-Build Turbocharged NVIDIA GB200 Cluster with 36K Blackwell GPUs in Partnership with Hypertec Cloud." Together AI Blog, November 2024.
- ^1 ^2"Introducing Together Code Sandbox & Together Code Interpreter." Together AI Blog.
- ^1 ^2"Fireworks.ai vs Together.ai vs Replicate vs Anyscale - Comparison." Oden, 2025. (Link dead as of November 2026; no archived copy.)
- ^1 ^2 ^3 ^4 ^5 ^6"Together AI Raises $305M Series B to Scale AI Acceleration Cloud for Open Source and Enterprise AI." PR Newswire, February 20, 2025.
- ^1 ^2 ^3"Together AI Acquires Refuel.ai to Accelerate Development of Production-Grade AI Applications." PR Newswire, May 15, 2025.
- ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10"Pricing | Together AI." Together AI, official pricing page, accessed September 28, 2026.
- ^Together AI. "Together AI official newsroom media resources". Together AI newsroom identifying media-use permission. Original PNG extracted unchanged from the official newsroom logo-suite ZIP.
- ^1 ^2 ^3 ^4 ^5 ^6"Neocloud Together AI raises $800M, leaps to $8.3B valuation." TechCrunch, July 1, 2026.
- ^1 ^2"Latham & Watkins Advises Together AI in Series C." Latham & Watkins, July 2026.
- ^"Together Post-Training: frontier-grade model shaping, the way that fits you." Together AI Blog, October 2026.
- ^1 ^2 ^3"Together AI Announces Business And Product Milestones At First AI Native Conference." PR Newswire via CIO Influence, March 6, 2026.
- ^"Together AI Achieves 90% Faster BF16 Training with NVIDIA Blackwell Platform and Together Kernel Collection." Together AI Blog, February 13, 2025.
- ^"Together Inference Engine 2.0." Together AI Blog, July 18, 2024.
- ^"Together raises $20M to build open source generative AI models." TechCrunch, May 15, 2023.
- ^1 ^2"Together AI expands fine-tuning service with tool calling, reasoning, and vision support." Together AI Blog, March 18, 2026.
- ^"Introducing the Amazon Bedrock AgentCore Code Interpreter." AWS Machine Learning Blog, 2025.
Improve this article
Add missing citations, update stale details, or write a clearer explanation. Anyone can edit: changes go live when you publish, and every edit is checked by an editor.
16 revisions · v16 · 6,463 words · full history
Fact-checks are independent of edits: a reviewer re-verifies the article against its sources and stamps the date. How we verify
Research and drafting on this wiki are AI-assisted, under named human editorial standards. How AI is used here
Reviewer note: Independent fact-check of the full article against its cited sources (170 claims, 36 sources fetched); 8 material + 21 minor fixed
Cite this page: AI Wiki. "Together AI." aiwiki.ai, updated 29 Oct 2026, fact-checked 1 Nov 2026. CC BY 4.0. https://aiwiki.ai/wiki/together_ai


