Frontier AI Spend Holds Steady Even as Open-Source Use Surges

Platform data from Vercel and OpenRouter shows open-source models winning on token volume, but Anthropic and other frontier labs still capturing the majority of actual dollars spent.

Despite a sharp rise in open-source model usage, frontier AI providers like Anthropic are holding on to the bulk of developer spending — at least for now, according to platform data cited in TechCrunch reporting published July 7 [1].

The data matters because it tests a widely held assumption: that cheaper open-source models will eventually cannibalize the revenue of expensive frontier labs. So far, the numbers don’t support that conclusion [1].

What the platform data shows

On Vercel’s AI gateway — which routes traffic across multiple model providers — DeepSeek has surged to first place by token volume, processing just over a third of all tokens passing through the platform in the past week [1]. Z.ai, the lab behind the GLM-5.2 model, jumped to fourth place over the same period [1].

But scroll to total spend and the picture flips: Anthropic still accounts for more than half of all AI spending on the Vercel platform [1]. That share has dipped slightly over the past month, partly because of Anthropic’s own price increases, but not by much [1].

OpenRouter, which serves a broader and somewhat less enterprise-focused segment of the market, tells a similar story [1]. DeepSeek V4 Flash leads on raw usage at 5.3 trillion tokens processed weekly, while the most popular frontier model, Opus 4.8, handles just over 2 trillion [1]. OpenRouter does not rank models by total spend, but it logs the average token cost for Opus 4.8 at roughly 23 times that of V4 Flash — $1.37 per million tokens versus 6 cents — suggesting Opus is still capturing the larger share of dollars [1].

A two-phase life cycle, not a zero-sum race

The framing comes from Decagon chief executive Jesse Zhang, who published a post titled “Everyone is wrong about open source AI in the enterprise” on Monday [1]. Zhang argues that frontier and open-source models are not direct competitors but rather two stages of the same deployment cycle: expensive frontier models prove out new use cases, which then get handed off to cheaper open-source alternatives once they mature [1].

The result, in his telling, is that as older use cases migrate to lighter models, new ones keep emerging at the frontier — keeping overall spend on top-tier models roughly flat [1]. As Zhang puts it, “The frontier labs will keep owning discovery. Open source will increasingly own production.” [1]

TechCrunch notes that Zhang offers limited data of his own to back the claim, but points to the Vercel and OpenRouter figures as broadly consistent with his thesis [1].

One more variable: Nvidia’s Nemotron

The current figures do not yet capture Nvidia’s Nemotron model, which TechCrunch reports is positioned to jump to the front of the usage rankings, citing Nvidia’s industry relationships and the model’s “extreme adaptability” [1].

A second explanation for frontier labs’ resilience, beyond Zhang’s life-cycle theory, is that many production use cases remain too complex to be fully replaced by cheaper alternatives — keeping a floor under demand for the most capable models [1].


Sources

  1. TechCrunch — Why the rise of open source AI isn’t hurting Anthropic … yet

This article was drafted with AI from the cited sources and checked against them before publication. Spot an error? Let us know.

The AI company unveiled a scientific workbench for researchers and said it will pursue drug discovery for neglected diseases — a rare move for a frontier model maker.

Anthropic announced Claude Science, a new AI workbench for scientists, and said it plans to develop its own drugs, making it one of the first major frontier AI companies to pursue drug development directly [1].

The dual announcement came at an event called “The Briefing: AI for Science” earlier this week [1]. Claude Science is designed to pull fragmented tools and datasets into a single environment and generate figures and visuals for researchers [1]. Anthropic framed the launch around what it describes as AI’s potential to “dramatically accelerate the pace of scientific discovery and the development of healthcare interventions,” and pointed to an existing list of biotech and pharma customers already using Claude [1].

Going Beyond Software

The more unusual move is Anthropic’s stated intent to discover drugs itself. Eric Kauderer-Abrams, the company’s head of life sciences, said Anthropic will focus on treatments for “neglected” diseases [1]. The company did not respond to The Verge’s requests for additional detail — including which diseases it plans to target first, and whether it would partner with other organizations for lab work, animal testing, clinical trials, or manufacturing [1].

That puts Anthropic in the position of selling software to drugmakers who could also become its competitors [1]. It joins a field that already includes AI-first drug companies such as Insilico and Isomorphic Labs — a spinout from Google DeepMind — as well as biotech startups and large pharmaceutical companies building or acquiring AI tools [1]. OpenAI, Amazon, and Google also have their own life sciences tools and platforms [1].

To support the effort, Anthropic has been hiring biologists and building its own wet labs over the past year, and had several open life sciences roles at the time of publication [1]. Namshik Han, a professor at the University of Cambridge and co-founder of AI biotech startup CardiaTec, told The Verge that Anthropic has been “actively recruiting” in the area and that he believes the company has hired candidates away from Big Pharma and prestigious academic institutions [1].

What AI Can — and Cannot — Do in Drug Discovery

Experts who spoke with The Verge were measured about what AI can realistically deliver. Han said AI is applied at “every single stage of drug discovery,” from finding new compounds to supporting clinical trials and manufacturing [1]. Matthew Todd, a professor of drug discovery at University College London, called “AI drug discovery” a “catchall phrase” given how broadly the technology is used across the field [1].

AI can already help generate candidate molecules and speed up early research, Han said, for example by suggesting compounds that could interact with cell receptors already known to be involved in a particular disease [1]. Todd said it is useful for “road testing” new drug ideas [1]. Given Anthropic’s work on frontier models, the company would presumably use generative AI to search across large chemical and biological datasets and help researchers surface connections that would otherwise be slow to find [1].

But the path from AI-generated candidate to approved medicine remains long. Todd said the field is “a long way off” from an AI-designed drug receiving regulatory approval for human use, and stressed that the discovery process would not run autonomously — human input and supervision would be required throughout [1]. Both Todd and Han flagged a shortage of publicly available, high-quality experimental data on how chemicals behave in the body as a potential bottleneck, even in well-studied areas of biology [1].

Frank von Delft, a professor of structural chemical biology at the University of Oxford and head of protein crystallography at the Oxford Centre for Medicines Discovery, said AI models “haven’t yet come close to making experiments unnecessary” [1]. Drug candidates still need real-world testing for efficacy, toxicity, and practical properties such as safe preparation, storage, and delivery — work that requires skilled staff, significant funding, and time [1]. If Anthropic wants to develop a drug, von Delft said, it is “going to have to spend a lot on experiments” [1].

No AI-designed drug has yet completed clinical trials and received approval from the U.S. Food and Drug Administration (FDA) to reach market, though some AI-developed candidates have entered trials [1]. Todd noted there is “always a big lag time” with medicine testing, adding: “It takes time to show experimentally that something’s safe” [1]. Any payoff from Anthropic’s drug development effort is likely at least a decade away, given typical clinical trial timelines [1].


Sources

  1. The Verge — Anthropic wants to develop its own drugs

This article was drafted with AI from the cited sources and checked against them before publication. Spot an error? Let us know.

After a three-week regulatory shutdown triggered by an Amazon security report, Claude Fable 5 is back online worldwide — but with new government commitments, a safety-layer trade-off, and Mythos 5 still restricted.

Anthropic restored global access to Claude Fable 5 on July 1, 2026, after the U.S. Department of Commerce withdrew the emergency export-control order it had issued on June 12 that forced the company to pull the model from all customers worldwide. [1]

The official Claude account on X announced the model’s return at 3:31 PM ET on July 1. [1] For organizations using cloud hyperscalers, Anthropic said it is working to re-enable access on Amazon Web Services, Google Cloud, and Microsoft Foundry “as quickly as possible,” though VentureBeat was unable to confirm restoration on those platforms at time of publication. [1]

Fable 5’s cybersecurity-focused counterpart, Claude Mythos 5, remains in a more restricted state. Commerce Secretary Howard Lutnick’s June 30 letter to Anthropic executive Tom Brown stated that a license is no longer required for the export, reexport, or in-country transfer of either model. [1] However, Anthropic’s own redeployment post says Mythos 5 access has been restored only for “a set of US organizations,” following government approval on June 26, with broader access being coordinated through its opt-in cybersecurity testing program, Project Glasswing. [1]

How the Crisis Unfolded

Anthropic launched Fable 5 and Mythos 5 on June 9, 2026, with early enterprise results drawing attention — Stripe reported that Fable 5 compressed a codebase-wide migration across a 50-million-line Ruby infrastructure into a single day, a project estimated to take a team more than two months by hand. [1]

Three days later, on June 12 at 5:21 PM ET, the U.S. government issued an export-control directive citing national security authorities, banning access to the models by any foreign national whether inside or outside the United States. [1] Because Anthropic lacked real-time mechanisms to verify user nationality at the application programming interface (API) layer, it was forced to suspend access for all customers to ensure compliance. [1] Access to all other Anthropic models was not affected. [1]

The technical catalyst was a report by Amazon researchers describing a method for bypassing Fable 5’s safeguards — a notable irony given Amazon is one of Anthropic’s largest backers, having invested $8 billion in the company. [1] According to Anthropic, the technique prompted Fable 5 to identify software vulnerabilities, and in one case the model produced code demonstrating how a relevant vulnerability could be exploited. [1]

Anthropic countered that the exploit did not tap into unique capabilities exclusive to Mythos 5, noting that its own testing found other models — including Claude Opus 4.8, OpenAI’s GPT-5.5, and Moonshot’s Kimi K2.7 — could identify the same vulnerabilities, and that every model it tested could produce the same exploit demonstration. [1]

The Technical Fix and Its Operational Cost

To break the regulatory deadlock, Anthropic developed an improved automated safety classifier specifically trained to catch and neutralize the Amazon technique. [1] Tested by the Commerce Department’s Center for AI Standards and Innovation (CAISI), the updated classifier successfully halts that specific technique in more than 99% of cases. [1]

Anthropic explicitly warns enterprise clients that this safety enforcement comes at an operational cost: because the new classifiers require an expanded “safety margin” to catch ambiguous edge cases, benign coding and debugging requests may be flagged more often. [1] When a prompt is blocked by the safety layer, the active session automatically downgrades, routing the request to Opus 4.8. [1] Thariq Shihipar, a Member of Technical Staff at Anthropic working on Claude Code, posted on X that Anthropic is “continuing to refine these safeguards to better distinguish genuine misuse from legitimate requests and reduce false positives.” [1]

The Political Dimension

The breakthrough was as much political as technical. According to WIRED, Anthropic initially argued that the administration’s security concerns were overblown and that no frontier model provider could guarantee zero jailbreaks — an argument that frustrated the administration. [1] In recent weeks, Anthropic shifted its approach, focusing on building stronger safeguards and satisfying the government’s operational concerns rather than debating the theoretical limits of model safety. [1]

WIRED reported that Anthropic CEO Dario Amodei was recently replaced in government meetings by Brown, whom officials liked more personally. [1] Under Brown’s guidance, Anthropic committed to a set of obligations outlined in the Commerce letter, including proactively detecting and addressing security risks, working with the U.S. government on protocols and standards for Mythos, Fable, and future models, and informing the government of malicious activity. [1] The Commerce Department explicitly reserved the right to re-evaluate these permissions and re-impose license requirements if circumstances change or if Anthropic fails to meet its commitments. [1]

Pricing and Temporary Incentives

Fable 5 and Mythos 5 are priced at $10.00 per million input tokens and $50.00 per million output tokens, making them the most expensive frontier models globally by that measure. [1] For comparison, OpenAI’s GPT-5.5 is priced at $5.00 per million input tokens and $30.00 per million output tokens, and Google’s Gemini 3.1 Pro Preview runs at $2.00 per million input tokens and $12.00 per million output tokens for contexts up to 200,000 tokens. [1]

To offset disruption from the export-control saga, Anthropic is running a temporary rollout through July 7 in which Fable 5 usage is included at no added cost for up to 50% of a user’s weekly tier allowance for Pro, Max, Team, and select Enterprise subscriptions. [1] After July 7, Fable 5 will move to usage credits for those plans. [1] For standard Enterprise seats, there is no included Fable 5 allowance; all usage is billed through credits, and the model will not function for those users unless credits are enabled. [1]

Enterprise users should also note Anthropic’s mandatory 30-day data retention requirement for covered models: prompts and model completions are retained for at least 30 days by default and then automatically deleted, except when they are part of a safety investigation or must be kept for legal reasons. [1] Organizations in regulated industries — financial services, healthcare, and legal — will need to evaluate whether this window complies with their data privacy mandates. [1]

Broader Industry Implications

The Fable 5 episode is not isolated. OpenAI’s newest models — GPT-5.6 Sol, Terra, and Luna — are currently in limited preview for a small group of trusted partners after the U.S. government requested a staggered rollout following a preview of the models and their capabilities. [1] OpenAI stated it “don’t[s] believe this kind of government access process should become the long-term default,” arguing it “keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.” [1]

The regulatory backdrop is a June 2, 2026, executive order signed by President Donald J. Trump directing federal agencies to collaborate on a 30-day process for benchmarking and assessing the capabilities of new AI models before wide release. [1]

Critics of the original Fable 5 export control argued the U.S. risked damaging its own industry. Former Facebook security chief Alex Stamos called the Fable restriction a “huge own goal for the US,” warning that security companies could be driven toward Chinese models. [1] The episode has accelerated interest in model-agnostic fallback architectures, with enterprise architects exploring proxy layers that can dynamically reroute production pipelines from proprietary APIs to locally hosted, open-weights alternatives. [1]


Sources

  1. VentureBeat — Anthropic is bringing back Claude Fable 5 globally after US lifts export control order — where can enterprises access it?

This article was drafted with AI from the cited sources and checked against them before publication. Spot an error? Let us know.