Baseten raises $1.5 billion at $13 billion valuation as AI inference demand explodes

Baseten raises $1.5bn at up to $13bn for AI inference

The inference startup is raising at two valuations at once, $11bn for some investors and $13bn for others, as open-source models pile pressure on OpenAI and Anthropic. Baseten is finalising a $1.5bn funding round that values the company at up to $13bn. The structure is almost as notable as the size. The round is dual-tiered, with some investors buying in at an $11bn valuation and others at $13bn, according to the company via the Wall Street Journal. It is a tactic a number of AI startups have used lately to lift a headline number. The financing is co-led by Altimeter Capital, Conviction, Spark Capital, Sands Capital, and Wellington Management. Baseten sells the unglamorous part of the AI stack. The San Francisco company, founded in 2019, provides the software and computing capacity businesses need to run inference, the step where a trained model actually answers a query. It rents capacity from around 20 cloud providers and layers its own inference software on top, so customers can deploy and fine-tune models on their own data. Much of that work runs on cheaper, open-source models rather than the frontier systems from OpenAI or Anthropic. That is the whole bet. "Open-source models are getting very, very good," chief executive Tuhin Srivastava told the WSJ, adding that customers increasingly mix open and proprietary models depending on how hard the task is. A fast re-rating Investors are paying up for that thesis at speed. Baseten was valued at $5bn as recently as January, when it raised a $300mn round backed by IVP, CapitalG, and Nvidia. The new deal more than doubles that in roughly five months. As recently as September 2025, the company was worth about $2.15bn. The pull is demand. Customers including Cursor, Mercor, and OpenEvidence use Baseten to serve models, and the company says some have cut costs sharply by shifting workloads to open-source options, with one reportedly running a task at about 30 per cent of the cost of a proprietary alternative. The picks-and-shovels trade The raise lands in the middle of an AI inference gold rush, where the layer that makes models run efficiently has become some of the most sought-after infrastructure in the industry. Cerebras, an inference-chip maker, went public earlier this year, and Fireworks AI is chasing the same model-serving market. It also lands in the middle of a price war. Many of the most-used open-source models now come from China, including DeepSeek and Moonshot AI, and even Nvidia has released an open family called Nemotron. With OpenAI reportedly weighing steep price cuts to keep up, that pressure is bad news for the frontier labs and, in theory, good news for whoever helps companies run the cheaper options. The risk sits in the same place as the opportunity. Baseten is being valued as a winner of a cost collapse, even as that collapse squeezes margins across the field. A $13bn tag on a company at the centre of a price war is, itself, a bet that the shovels outlast the gold.

AI inference provider Baseten reportedly raising $1.5B in funding

AI inference provider Baseten reportedly raising $1.5B in funding Baseten Inc., a startup with a platform for running artificial intelligence inference workloads, is raising $1.5 billion in funding. The Wall Street Journal reported today that Altimeter Capital, Conviction, Spark Capital, Sands Capital and Wellington Management are co-leading the deal. It's unclear whether there are additional participants. Some of the investors are buying shares at a $11 billion valuation while the other backers' term sheets specify a $13 billion valuation. Setting up a cloud-based inference cluster involves a significant amount of work. Developers have to provision graphics cards, configure them, link them together and install a large number of software tools. Baseten provides a platform that automates the workflow. The software is available as a managed service and as a standalone application that companies can deploy in their public cloud environments. Baseten's platform is powered by three core modules the company calls inference engines. They optimize the performance of customers' AI models and collect data about technical issues. The first inference engine, BIS-LLM, is designed power large language models with a mixture of experts architecture. A mixture of experts LLM comprises multiple neural networks that are each geared towards different tasks. BIS-LLM improves the efficiency of such models by optimizing their KV cache, a data structure that stores information necessary for inference. When a model's token usage increases, BIS-LLM automatically provisions more hardware. Baseten's second inference engine is called Engine-Builder-LLM. It's optimized for dense LLMs, which are models that comprise a monolithic collection of artificial neurons rather than multiple neural networks. AI models usually generate output one token at a time. Engine-Builder-LLM uses a technology called lookahead decoding to generate multiple tokens at once, which speeds up processing. Baseten's third core inference engine, BEI, is geared towards simpler AI models. It can power embedding models, which turn raw data into a format that LLMs understand, as well as data classification and search models. Baseten uses a software module called MCM to spread inference workloads across multiple public clouds. If one of the clouds experiences an outage, MCM reroutes prompts to the platforms that are still online. According to Baseten, the technology's ability to switch providers is also handy when a company's main public cloud has a shortage of graphics cards. The platform provides out of the box support for several dozen open-source AI models. Additionally, customers can deploy custom algorithms using a tool called Truss. It automates the task of packaging an LLM into a Baseten-compatible format. Baseten can not only perform inference with custom LLMs but also train them. According to the company, its platform includes a backup feature that periodically saves copies of a neural network while it's being trained. If a technical issue crops up, developers can restore the most recent backup copy instead of starting the training workflow from scratch. Baseten's funding comes less than six months after its previous raise. The $300 million investment included contributions from Nvidia Corp. and CapitalG, Alphabet Inc.'s growth-stage startup investment arm.

Baseten Nears $1.5 Billion Funding Round as Inference Demand Surges | PYMNTS.com

The round will have a dual-tiered structure in which some participants will invest at an $11 billion valuation and others at a $13 billion valuation, according to the report. Baseten's latest round reflects the growing demand for inference as companies look to use open-source models as a much-lower-cost alternative to increasingly expensive closed-source models, the report said. Companies can download open-source models at no cost, modify them to meet their needs, and turn to a company like Baseten for the inference software layer and computing capacity. Baseten sources computing power from 20 different cloud providers, per the report. Baseten Co-Founder and CEO Tuhin Srivastava told the WSJ: "At the highest level, what's happening generally in the market is that the open-source models are getting very, very good. And as open-source gets better, we are growing with it." It was reported in May that Baseten was holding talks with investors to raise $1 billion at an $11 billion valuation and that some investors were making offers valuing the company at around $15 billion. That report said that at the end of the first quarter, Baseten's annualized revenue came to around $600 million, compared to $200 million at the beginning of the quarter. The report attributed the growth to an explosion of apps using open-source AI models, which has been a boon for Baseten and other inference providers. Basten was valued at $5 billion in a January Series E funding round in which it raised $300 million. That round followed a September 2025 Series D round in which it was valued at $2.15 billion and raised $150 million. Following the September round, Srivistava told PYMNTS CEO Karen Webster in an interview: "For us, inference is one part of AI infrastructure. Beyond that, there's training, evaluation, fine-tuning. We really want to own that entire loop. We want to build the next AWS for inference." When announcing the Series E in a January blog post, the company's co-founders wrote that they started Baseten with the belief that inference would become the link between AI's promise and its real-world impact. They added that the demand for inference is accelerating. "As AI becomes embedded in every product and workflow, inference isn't just growing -- it's becoming one of the largest markets ever created," they wrote. "We're building the software that powers it."