Unlike proprietary closed-source AI systems, Inkling is an open-weight model, allowing developers and enterprises to download, run and customise the underlying model according to their specific requirements.

The launch marks the first general-purpose AI model released by Thinking Machines, the San Francisco-based startup founded in 2025 by former OpenAI Chief Technology Officer Mira Murati.

Previously, the company introduced its first product, Tinker, in October last year. The platform enables users to customise AI models, and the company has now made Inkling available through Tinker as well as other developer platforms.

Inkling features 975 billion parameters—the variables that determine how an AI model processes and understands information—making it one of the largest open-weight models introduced to date.

The launch comes as the Western open-source AI ecosystem continues to trail China’s rapidly growing AI landscape. The gap widened after Meta shifted its strategy toward proprietary AI development following the underwhelming reception of its Llama 4 open model.

As a result, many businesses have increasingly adopted Chinese open-source AI models as cost-effective alternatives to expensive proprietary systems.

For example, hedge fund Bridgewater Associates used Tinker to build a customised version of Qwen, Alibaba’s AI model, which the company said outperformed several leading proprietary models while operating at lower costs.

To demonstrate Inkling’s capabilities, Thinking Machines published benchmark results comparing the model against proprietary AI systems developed by Anthropic, Google, and OpenAI, as well as several leading open-weight models, most of which originate from Chinese AI companies.

Although the competing models continue to lead overall performance, Inkling delivered competitive benchmark results, particularly in AI agent-related tasks, highlighting its potential to attract developers and enterprise customers seeking flexible and customisable AI solutions.

With the launch of Inkling, the AI startup aims to strengthen the Western open-weight AI ecosystem while expanding its presence in the increasingly competitive global artificial intelligence market.

Private equity firm Creaegis led the funding round. Meanwhile, new investors MNI Ventures-Claypond and Sentinel Global joined existing backers Khosla Ventures, SoftBank Vision Fund 2, Lightspeed, and Y Combinator in the investment. With this latest round, Emergent’s total funding has reached $230 million. Earlier this year, the startup raised $70 million in a Series B round at a $300 million valuation.

The AI coding sector continues to attract strong investor interest as startups including Lovable, Replit, and Cursor secure significant funding to develop tools that help developers build software more efficiently. At the same time, leading AI companies such as OpenAI and Anthropic have expanded their focus on coding solutions.

Amid rising competition, Emergent aims to differentiate itself by serving entrepreneurs, startups, and small and medium-sized businesses that continue to rely on emails, spreadsheets, and messaging platforms to manage their operations.

“Our thesis has always been to build a production-grade application for serious builders,” Emergent co-founder and chief executive Mukund Jha said in an interview. “So you’re basically getting an engineering team in a box.”

The company has achieved an annual revenue run rate of $120 million, reflecting 70% growth over the past four months, while its platform now serves more than 200,000 paying customers. Mukund Jha founded Emergent alongside his brother Madhav Jha (CTO) in June 2025.

Today, customers across multiple industries use Emergent’s platform to build business applications. Its client base includes trucking companies developing shipment tracking systems, manufacturing businesses creating enterprise resource planning (ERP) software, construction firms building operational platforms, and property management companies developing internal customer management solutions.

North America contributes nearly one-third of the company’s revenue, while Europe generates another one-third. The remaining revenue comes from other international markets, with India accounting for approximately 8% to 9%, according to Mukund Jha.

Emergent’s business strategy places it in direct competition with Replit, which Jha identified as the company’s closest rival. However, he differentiated Emergent from developer-focused AI coding tools such as Anthropic’s Claude Code, OpenAI’s Codex, and Cursor, explaining that non-technical users require a platform capable of handling deployment, hosting, testing, and debugging alongside software development.

At the same time, Jha acknowledged that the company continues to improve one important aspect of its platform.

He noted that many AI-generated websites currently share similar visual designs, making design differentiation an area requiring further innovation.

Looking ahead, Emergent will deploy the fresh capital to accelerate product development and research. The company also plans to improve the success rate of applications built on its platform while enhancing its core AI agent workflows.

Additionally, Emergent is developing support for more advanced AI applications, including those powered by local and open-source models. The startup also intends to strengthen its global go-to-market strategy and expand its international customer base.

Furthermore, the company is evaluating plans to establish a European office as customer demand across the region continues to grow.

Currently, Emergent employs around 200 people, with the majority based in Bengaluru and a smaller team operating from San Francisco. By the end of the year, the startup plans to expand its San Francisco workforce by 30 to 40 employees, further supporting its international growth strategy.

PixVerse completed the initial tranche of its Series C funding in March, led by CDH Investments. Although the company did not disclose the amount raised at the time, it had reported that the round was valued at around $300 million.

The latest extension attracted investments from Alibaba, Lollapalooza Capital, Ivy Capital, Grand Mount Capital, Eastern Bell Capital, Mirae Asset, BlueFocus, and CloudAlpha. Existing investors iGlobe Partners and Lion X Ventures, the venture arm of OCBC, also participated in the round.

Founded in 2023 by Wang Changhu and Jaden Xie, PixVerse has rapidly emerged as a major player in the AI video-generation market. Before launching the company, Changhu worked on computer vision technologies at ByteDance, while Xie served as an executive director at investment firm Lighthouse Capital.

PixVerse currently offers multiple AI models tailored to different user segments. Its V-Series supports consumer applications and APIs, while the C-Series targets professional filmmaking and commercial production workflows. Earlier this year, the company also introduced its R-Series world models designed for game development and virtual world creation.

The platform enables users to generate videos with integrated audio in resolutions of up to 4K. According to the company, its consumer platform has attracted more than 150 million registered users and over 15 million monthly active users. While PixVerse declined to disclose the number of paying subscribers, it currently charges $4.80 per minute for image-to-video generation.

Discussing the competitive landscape, Jaden Xie, Co-founder, PixVerse, said only a handful of companies currently possess the technology required to produce high-quality AI-generated videos.

“OpenAI exited the business when they shut down Sora 2. Other companies like Meta and Tencent are not able to create high-quality video models. So there are only a few companies that can meet the quality bar,” he said.

Xie also said that PixVerse sees significant opportunities in both consumer and enterprise markets, as individuals increasingly create AI-generated entertainment content while businesses adopt AI video for marketing, training and creative production.

However, the company believes its competitive edge extends beyond model quality. Instead, PixVerse attributes its advantage to its proprietary data-labelling capabilities.

“We think the key difference is not in data but in how you label it, because data is available everywhere. My co-founder worked at ByteDance, where he built core visual understanding technology behind TikTok using AI. Using this tech, TikTok was able to label data accurately and build a strong recommendation algorithm. This experience comes in handy when building a video-generation platform,” Xie said.

Looking ahead, PixVerse plans to accelerate its global enterprise expansion. The startup has already partnered with Alibaba to deploy its AI video-generation capabilities across enterprise use cases.

Additionally, the company intends to launch a new V-Series video-generation model and introduce the next version of its world model later this year. PixVerse currently employs around 150 people across offices in Singapore, Beijing and Shanghai and plans to recruit more AI researchers and go-to-market professionals using the newly raised capital.

Despite its rapid growth, PixVerse faces intensifying competition in the AI video-generation market. Rivals include ByteDance’s Seedance; Video Rebirth, founded by former Tencent AI executive Dr. Wei Liu; and Kling AI, as well as Western competitors such as Midjourney, Runway and Luma. Meanwhile, several companies, including startups led by Yann LeCun and Fei-Fei Li, are also developing next-generation world models.

With its latest funding round, PixVerse has significantly strengthened its financial position as it competes in the fast-growing AI video-generation market. The company now aims to expand globally, accelerate product innovation and establish itself as a leading provider of AI-powered video and world model technologies.

The report stated that Crusoe currently supplies AI computing power to leading technology companies, including Meta and Oracle. Consequently, the startup has positioned itself among the fastest-growing providers of specialized AI infrastructure. As generative AI (GenAI) adoption expands worldwide, major technology companies continue to invest billions of dollars in advanced data centers to support increasingly complex computing workloads.

Meanwhile, Crusoe continues to negotiate with prospective investors, and the company has not finalized its valuation. However, investors reportedly expect the startup’s enterprise value, including the proposed investment, to reach approximately $30 billion.

Previously, Crusoe secured $1.38 billion during its Series E funding round in 2025, achieving a valuation exceeding $10 billion. Valor Equity Partners and Mubadala Capital co-led that investment round, reinforcing investor confidence in the company’s long-term AI infrastructure strategy.

Crusoe originally launched in 2018 as a cryptocurrency-focused business. However, the company later shifted its business model and now develops AI infrastructure solutions that power next-generation artificial intelligence applications. As a result, Crusoe has become one of the emerging neocloud providers that specialize in AI cloud services, GPU computing, and hyperscale data center operations.

Furthermore, Crusoe announced in June that it had secured contracts covering 4.9 gigawatts of computing capacity. In addition, the company revealed that its overall project pipeline had expanded to more than 40 gigawatts, demonstrating strong customer demand and a rapidly growing portfolio of AI infrastructure projects.

Crusoe’s proposed $3 billion funding round underscores the accelerating investment in AI data centers, cloud infrastructure, and high-performance computing. If the company completes the fundraising at the expected valuation, Crusoe will further strengthen its position as a leading AI infrastructure provider while capitalizing on the rapidly expanding global demand for generative AI computing capacity.

“The adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce,” Oracle said in an annual financial regulatory filing.

The technology giant reported that its global full-time employee count declined to 141,000 as of May 31, the end of its fiscal year, compared with 162,000 employees a year earlier. As a result of the workforce restructuring, Oracle incurred around US$1.8 billion in related costs.

The workforce reduction comes as Oracle faces mounting financial pressure from its aggressive investment in AI infrastructure. The company continues to expand its network of AI-focused data centres to support major clients, including OpenAI. Earlier this year, reports indicated that Oracle had begun eliminating thousands of positions as part of broader cost-saving initiatives, although the company had not publicly disclosed the full extent of the cuts at that time.

According to the filing, the company employed approximately 49,000 workers in the United States at the end of May, while around 92,000 employees worked across its international operations.

Oracle’s current workforce is now slightly smaller than it was before the company acquired Cerner in 2022. The US$28 billion acquisition significantly expanded Oracle’s employee base, adding thousands of workers, many of whom were located near Cerner’s headquarters in the Kansas City metropolitan area.

The latest workforce reduction highlights how major technology companies continue to use artificial intelligence to streamline operations while simultaneously investing billions of dollars in AI infrastructure and cloud computing capabilities.

Following the report, Qualcomm’s shares slipped approximately 1% in extended trading, reflecting investor reaction to the potential transaction.

According to the report, negotiations remain ongoing, and both the valuation and deal structure could still change. Additionally, the discussions could ultimately fail to result in an agreement. The report noted that it remains unclear whether the proposed purchase price would include performance-based milestone payments, a mechanism that acquirers have frequently used in previous semiconductor startup acquisitions.

The potential acquisition aligns with Qualcomm’s broader strategy to diversify beyond its traditional smartphone chip business. As one of the world’s largest suppliers of mobile processors, the company has increasingly focused on expanding its presence in high-growth technology segments, including artificial intelligence, data center processors, and autonomous vehicle chips.

Founded in 2016, Tenstorrent has emerged as a notable player in the AI semiconductor industry. The company develops specialized accelerators designed for training artificial intelligence models and running AI applications, positioning itself within one of the fastest-growing areas of the global technology market.

Tenstorrent operates under the leadership of Jim Keller, a former Apple chip designer who also oversaw Tesla’s efforts to develop chips for autonomous driving systems. Under his leadership, the startup has attracted industry attention for its AI-focused hardware innovations and ambitions to compete in the rapidly evolving AI infrastructure space.

If completed, the acquisition would represent one of the largest deals in the AI chip sector and could significantly strengthen Qualcomm’s position in the race to capitalize on the growing demand for artificial intelligence computing power. Furthermore, the move would accelerate Qualcomm’s efforts to build a stronger foothold in data center and AI infrastructure markets as companies worldwide invest heavily in next-generation AI technologies.

As competition intensifies among semiconductor companies seeking leadership in AI hardware, a potential Tenstorrent acquisition could provide Qualcomm with advanced AI accelerator technology and experienced engineering talent, further enhancing its long-term growth strategy beyond smartphones.

Although the companies did not disclose the financial details of the partnership, a source familiar with the agreement said that the deal includes a fivefold expansion of Lovable’s Google Cloud footprint, including a significant increase in AI-related usage. As part of the collaboration, Lovable will also gain broader access to both Anthropic’s Claude models, which developers widely use for coding tasks, and Google’s Gemini family of AI models.

The Anthropic component of the agreement carries particular significance given Google’s growing investment in the AI company. In April, Google invested $10 billion in Anthropic through a combination of cash and compute credits. The technology giant also committed an additional $30 billion contingent upon Anthropic achieving specific performance milestones.

Google completed the investment at a valuation of approximately $350 billion. However, just one month later, Anthropic secured a massive $65 billion funding round that pushed its valuation close to $1 trillion. Consequently, the expanded partnership with Lovable could contribute toward Anthropic meeting those performance targets, as Lovable has emerged as one of Europe’s fastest-growing startups.

According to the company, Lovable surpassed $400 million in annualized revenue in February after adding $100 million in revenue within a single month. Remarkably, the company achieved this growth with a workforce of only 146 employees. Furthermore, Lovable claims that more than half of Fortune 500 companies use its platform in some capacity.

The new agreement also connects Lovable more deeply with Google’s broader enterprise ecosystem. Under the partnership, Lovable’s latest AI agent will become available through Google Cloud’s enterprise agent marketplace, the Gemini Enterprise Agent Gallery. Google and Lovable initially signaled this collaboration during Google’s major cloud conference in the United States earlier this year.

In addition, Lovable will integrate with Wiz, Google’s largest acquisition to date. Google completed the $32 billion acquisition of Wiz in March, approximately one year after announcing the deal. Through this integration, Wiz will help identify and remediate security vulnerabilities in real time across code generated by both human developers and AI agents.

Google believes that distributing Lovable’s AI agents through its marketplace will simplify enterprise procurement and billing processes. As a result, enterprise customers will gain easier access to Lovable’s solutions, potentially accelerating the startup’s customer acquisition efforts and strengthening its position in the enterprise AI market.

For Google, the strategic rationale behind the partnership remains straightforward. By supporting the continued growth of both Lovable and Anthropic while attracting enterprise customers with substantial technology budgets, Google can generate additional cloud revenue. This revenue becomes increasingly important as the company plans to invest between $180 billion and $190 billion in capital expenditures during the current year.

To help finance these ambitious investments, Google is already pursuing a record-breaking $85 billion equity sale. Nevertheless, the company still faces the challenge of securing substantial additional funding to support its long-term artificial intelligence, cloud computing, and infrastructure expansion strategies.

As competition intensifies across the global AI ecosystem, Google’s expanded collaboration with Lovable demonstrates how major technology companies are increasingly partnering with high-growth AI startups to drive enterprise adoption, strengthen cloud revenues, and accelerate innovation. Simultaneously, the partnership provides Lovable with access to advanced AI models, enterprise distribution channels, and enhanced security capabilities, positioning the company for continued rapid growth in the evolving artificial intelligence market.

The expanded partnership between Lovable and Google marks a significant milestone in the rapidly evolving AI and cloud computing landscape. By increasing its Google Cloud usage, gaining broader access to Claude and Gemini models, and integrating with Wiz, Lovable is strengthening its ability to serve enterprise customers at scale. Meanwhile, Google continues to reinforce its AI ecosystem by supporting high-growth startups that can drive cloud adoption and enterprise AI deployment. As both companies pursue aggressive growth strategies, this collaboration highlights the increasing importance of strategic alliances in shaping the future of artificial intelligence and enterprise technology.

Despite the strong earnings performance, the company’s shares remained largely flat in volatile extended trading after investors reacted to a sharp increase in operating expenses.

Demand for AI cloud infrastructure services from so-called “neocloud” providers such as CoreWeave and its peer Nebius has accelerated significantly as businesses race to secure computing power for generative AI models, machine learning systems, and enterprise AI applications.

CoreWeave reported total first-quarter revenue of $2.08 billion, exceeding analysts’ average estimate of $1.97 billion, according to data compiled by LSEG. However, the company’s operating expenses more than doubled year-on-year to $2.22 billion during the quarter.

The AI infrastructure business remains highly capital-intensive, and CoreWeave continues investing aggressively in expanding its data centre capacity to meet growing customer demand. The company’s expansion strategy requires billions of dollars in upfront infrastructure investments, particularly in advanced GPUs, AI servers, and cloud computing facilities.

A major competitive advantage for CoreWeave comes from its specialised AI infrastructure and close partnership with Nvidia, which provides the company with early and large-scale access to some of the world’s most sought-after AI chips and hardware technologies. Consequently, CoreWeave has become a preferred cloud provider for AI startups as well as enterprise customers seeking alternatives to capacity-constrained traditional cloud providers.

The company has recently secured several high-profile deals that further strengthen its position in the rapidly expanding AI cloud computing market. Over the past month, CoreWeave signed an expanded $21 billion agreement with Meta for additional cloud computing capacity. Additionally, the company entered into a $6 billion deal with trading firm Jane Street and secured another partnership agreement with Anthropic.

CoreWeave also reported a substantial increase in its revenue backlog, which reached $99.4 billion as of March 31, compared with $66.8 billion at the end of December. The expanding backlog highlights the growing global demand for AI infrastructure, cloud GPU services, and large-scale computing capacity required for next-generation artificial intelligence development.

CoreWeave’s latest financial results underscore the explosive growth in the AI infrastructure market as enterprises and AI developers continue investing heavily in cloud computing and high-performance AI systems. While rising operating costs reflect the capital-intensive nature of the business, CoreWeave’s expanding customer base, strategic partnerships, and massive revenue backlog position the company as a major player in the global AI cloud ecosystem.

The company linked the layoffs to what it described as an “agentic AI-first operating model” rather than employee performance or short-term cost-cutting measures. According to sources, co-founders and chief executives Matthew Prince and Michelle Zatlyn shared the announcement with employees through an internal message.

Cloudflare employed 5,156 full-time workers at the end of 2025, based on company filings. Additionally, the company expects to incur restructuring charges ranging between $140 million and $150 million, with most expenses expected during the second quarter of 2026.

In a company blog post, Prince and Zatlyn revealed that Cloudflare increased its internal use of AI tools by more than sixfold during the past three months. “The way we work at Cloudflare has fundamentally changed,” the executives wrote, adding that employees across departments already use “thousands of AI agent sessions each day.”

The leadership team stated that Cloudflare is now “reimagining every team and function” for what it called the “agentic AI era,” where AI systems increasingly handle software development, operations, finance, customer support, and internal workflows. Furthermore, the company clarified that the restructuring reflects a redesign of processes and operational roles rather than concerns related to productivity or immediate financial challenges.

Cloudflare also said the restructuring aims to help the organisation adapt to faster AI-assisted execution, automation, and accelerated product development cycles. According to The Wall Street Journal, the company rapidly expanded AI adoption internally in recent months, prompting major operational shifts across multiple departments.

The layoffs come despite strong financial performance from the company. For the first quarter, Cloudflare reported a 34% year-on-year increase in revenue, reaching $639.8 million. Meanwhile, adjusted earnings stood at 25 cents per share, exceeding Wall Street expectations. However, the company’s shares declined in extended trading after it issued second-quarter revenue guidance slightly below market forecasts. Cloudflare projected second-quarter revenue between $664 million and $665 million.

Prince described artificial intelligence as “a fundamental re-platforming of the Internet” and emphasized that the AI transition represents one of the biggest growth opportunities in the company’s history.

Cloudflare now joins a growing list of technology companies reshaping their workforce structures around artificial intelligence, automation, machine learning, and AI-powered software development. As enterprises increasingly integrate generative AI and intelligent automation into daily operations, many tech firms continue redesigning teams and workflows to align with the rapidly evolving AI economy.

Cloudflare’s workforce reduction signals the accelerating shift toward AI-driven business models across the global technology industry. While the company continues delivering strong financial growth, its restructuring highlights how artificial intelligence is transforming operational strategies, workforce planning, and the future of enterprise software development.

The venture investment arm of Meituan led the funding round, pushing Moonshot AI’s valuation to more than $20 billion, according to a statement released by financial advisor HF Capital, which advised several investors involved in the transaction. Additionally, the company reported that its annual recurring revenue crossed $200 million in April, driven primarily by subscriptions to its Kimi chatbot and AI model services.

A spokesperson for Moonshot AI did not respond to requests for comment. However, a representative from Meituan’s Long-Z Investments confirmed the firm’s participation in the round.

The latest investment demonstrates the rapid growth of Moonshot AI, which has increased its valuation more than fourfold within just a few months. Toward the end of last year, the Beijing-based startup raised $500 million at a valuation of $4.3 billion. Earlier this year, the company secured another $700 million at a $10 billion valuation. Subsequently, it was reported that the startup sought an additional $1 billion in funding through an expanded fundraising round.

Investors have increasingly directed capital toward a select group of Chinese AI startups that aim to compete globally in generative AI, large language models, AI chatbots, and enterprise AI services. Meanwhile, DeepSeek has reportedly begun raising external capital for the first time, attracting strong interest from state-backed investors at a potential valuation of up to $50 billion.

At the same time, Moonshot AI’s competitors, including MiniMax and Zhipu AI, have achieved valuations exceeding $30 billion following their strong market debuts in Hong Kong earlier this year. The developments reflect the accelerating momentum in China’s artificial intelligence ecosystem as companies race to develop globally competitive AI technologies.

Moonshot AI founder Yang Zhilin established the company after working at Meta Platforms and Google. The startup currently offers tiered subscription plans for its Kimi chatbot while also providing AI infrastructure and underlying model technology to enterprise customers.

As investor confidence in generative AI, enterprise AI solutions, and large language models continues to rise, Moonshot AI is rapidly emerging as a significant challenger to major Western AI companies in the global technology landscape.