Databricks has signed a term sheet for a Coatue-led financing at a $188 billion valuation, while unidentified sources put the round at $3 billion. The proposed capital would deepen its push into AI governance, data agents and operational databases, but the transaction remains open and the company supplied no new operating figures.
Databricks is seeking another private-market step-up before investors can compare the company with public peers on current, standardized results. The financing would give it more capital to compete across data infrastructure and enterprise AI, but a signed term sheet is not a closed round.
Databricks said in its July 16 announcement that existing investor Coatue is leading a strategic financing at a $188 billion valuation. New and existing investors are expected to join, and the company expects the round to close later this summer.
The company did not disclose how much it plans to raise. The $3 billion figure comes from people described as familiar with the matter, not from Databricks or Coatue. Coatue declined to comment when asked about that report, according to a separate account.
The reported price is 40% above the $134 billion valuation Databricks reached in December 2025. That earlier financing is an important comparison: Databricks announced it was raising more than $4 billion, while the later account said the completed fundraising totaled about $5 billion. A reported $3 billion follow-on would therefore extend a large recent capital build, not end a long financing drought.
In the December company release, Databricks said its third-quarter revenue run-rate had exceeded $4.8 billion, up more than 55% year over year. It put both AI products and data warehousing above $1 billion in revenue run-rate and said it had generated positive free cash flow over the preceding 12 months. The company also said Lakebase had thousands of customers, net retention exceeded 140%, and more than 700 customers were consuming at an annual revenue run-rate above $1 million.
Those figures are company-supplied snapshots from the previous round, not an update attached to the $188 billion proposal. The new announcement gives no current revenue, growth, retention or cash-flow figures. Nor does it disclose investor rights, pricing protections or whether any of the reported investment would buy existing shares. Without those terms and newer results, the valuation increase cannot be cleanly separated into operating progress and a higher price investors are willing to pay.
Databricks says the proceeds would accelerate three products:
The company also expects the capital to support future AI acquisitions and research. Ali Ghodsi, Databricks' co-founder and chief executive, framed the strategy as choosing the most suitable model for each task rather than using the most expensive model everywhere.
Unity AI Gateway can connect customers to models from OpenAI, Anthropic and Google, as well as open-weight alternatives, while tracking spending. That positions Databricks as an intermediary among model providers rather than a company dependent on one model vendor. It also makes the strategy dependent on enterprises trusting Databricks to mediate model choice, cost, data access and policy.
The economic problem is real, although the most striking figures originated with Databricks. In a June account, co-founder Patrick Wendell said he had seen companies move from little or no AI spending to accidentally spending tens of millions of dollars in a month. He said AI token costs were becoming a top-three expense for customers, behind salaries and other information-technology costs.
The gateway is designed to set spending limits, prevent runaway bills and recommend cheaper models for tasks that do not require the most expensive option. It will also monitor individual user sessions. That data could inform a decision to remove an employee's access or move the person to a cheaper model if usage is judged inefficient, the account said.
This is more than a billing feature. It gives employers a way to enforce AI-use policy at the user level and could redirect demand away from costly models. Wendell argued that the required data is narrow and tied directly to use of coding tools, but the same account noted that employee monitoring could be difficult to sell internally. Cost savings and worker acceptance are therefore separate deployment tests.
The financing may increase Databricks' capacity to build and buy products, but it does not make the product map unique. Snowflake says it has already launched Snowflake Intelligence for data agents and Snowflake Postgres for transactional workloads alongside analytics. It also describes a model-neutral strategy in which foundation-model providers deliver models within Snowflake's security perimeter.
The overlap extends beyond a single rival. In its latest annual filing, Snowflake lists AWS, Microsoft Azure and Google Cloud among competitors that can use control of their infrastructure, pricing and customer relationships to favor competing products. It also warns that frontier-model providers could expand into data storage and management. The contest for the enterprise AI control layer therefore includes data platforms, cloud infrastructure owners and potentially the model companies themselves.
The filing also shows why infrastructure economics matter. Snowflake said third-party cloud infrastructure represented about 71% of its cost of product revenue in its latest fiscal year, and it reported $2.68 billion of non-cancelable purchase commitments, much of them tied to cloud agreements. Those are Snowflake's economics, not Databricks', but they demonstrate that a software platform can mediate customer spending while still depending heavily on upstream infrastructure suppliers.
Direct financial comparisons require similar caution. Snowflake reported $4.7 billion of audited revenue for the fiscal year ended January 2026, up 29%, and a $1.3 billion net loss. Databricks' most recent figure in the retained sources is a company-reported revenue run-rate above $4.8 billion during a quarter. Annual revenue and a point-in-time run-rate are not like-for-like measures, and Databricks has not supplied the cost structure needed to compare margins or infrastructure exposure.
Databricks has long been viewed as a listing candidate. The report behind the $3 billion figure said the prospect of competing with SpaceX, OpenAI and Anthropic for public-market capital had led some private companies to reconsider their timing. It did not say Databricks had set or canceled an offering date.
Another private round would let Databricks fund product development, acquisitions and research without immediately submitting the $188 billion valuation to daily public-market pricing. It would not resolve when the company might list, and it would not establish that public investors would apply the same value.
The first evidence is transactional: a closed round at the announced valuation and the reported $3 billion size. Final disclosure would need to distinguish new capital from employee or investor liquidity and show any rights that affect the headline valuation.
The more consequential evidence is operating performance. Current revenue growth, free cash flow, retention and product-level adoption would show what changed after December. Customer results would need to demonstrate that Unity AI Gateway reduces costs without creating unacceptable monitoring or governance tradeoffs, and that Genie and Lakebase win workloads against products already offered by Snowflake, cloud providers and model vendors.
Until those facts arrive, the term sheet shows that investors are prepared to discuss a much higher private price. It does not yet show that Databricks has closed the financing or converted its broad AI platform strategy into results commensurate with $188 billion.
OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
RoboTTT extends a robot policy’s visuomotor context to 8,000 timesteps and raises its average rubric score from 42% to 79%, but the evidence comes from three author-run assembly tasks and the longest task was completed in only two of ten trials.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Nebius has arranged its first senior secured facility against an operating GPU deployment and one customer’s cash flows. The deal adds project-level debt to its expansion toolkit, but an unnamed customer and a larger rival financing limit what it proves about the rest of Nebius’s backlog.
Cursor says automatically running a repository-root git.exe on Windows does not meet its criteria for patching, while the researcher calls it an untrusted-search-path defect and separate research shows the same weakness across several competing AI coding tools.
Huawei publicly displayed a 16-cabinet Atlas 950 configuration rated at 1 EFLOPS in FP8, providing tangible evidence of its system-scale AI strategy while leaving price, power use and sustained workload performance undisclosed ahead of the full system's planned fourth-quarter release.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
UK testing places leading open-weight models four to seven months behind selected closed-model cyber results, yet longer attack chains, U.S. benchmarks and mixed cost comparisons show why that interval is a warning signal rather than a universal capability clock.
TSMC reached the top of its second-quarter revenue guidance and raised its 2026 outlook, but a one-off investment gain boosted profit growth while the company committed more capital to 2-nanometer production, advanced packaging and an undated Arizona expansion.
Moonshot AI's Kimi K3 added to a global technology selloff with near-frontier performance, but unreleased weights, mixed cost comparisons and a recommended 64-accelerator deployment leave its effect on chip demand unresolved.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
Two binding EU decisions require Google to give rival AI services comparable access to 11 Android features and offer eligible search competitors a restricted, anonymized dataset, but phased deadlines, certification, pricing and privacy safeguards leave the competitive effect unproven.
Netflix says generative AI workflows were used on roughly 300 titles in 2026, but its only quantified example tied to that disclosure covers 17 minutes and does not establish how much finished material, spending or labor changed across the slate.
Gemini 3.5 Pro missed its expected June rollout. An anonymous-source account says a late-June data change fell short of Google's coding goals, but Google has confirmed only partner testing—not the reported cause, a new date, or public results and pricing.
Google is bringing Gemini Omni editing and a reusable face-and-voice avatar to Vids, but the sharper distinction is account-level identity across Vids and Gemini rather than a new category of AI video; Vids already offered Veo generation and customizable avatars, while specialist rivals already sell digital presenters.
Intel plans to adopt Gemini Enterprise across engineering, supply-chain and corporate work while extending chip-simulation capacity into Google Cloud, but the companies supplied no rollout schedule, contract value or performance baseline.
The Navy has approved an immediate department-wide framework for turning data into operational effects, but the public rollout leaves budgets, owners, deadlines, performance measures and assurance rules to a promised implementation roadmap.
Runta has raised a $20 million seed round led by Andreessen Horowitz for an execution layer that governs agents at the operating-system and network levels, entering a market where cloud runtimes and sandbox providers already offer overlapping controls.