From Subsidies to Token-Based Pricing to Price Cuts: Is OpenAI Sparking a Price War? Is the Inflection Point for Token Economics Nearing?

marsbitPublié le 2026-06-11Dernière mise à jour le 2026-06-11

Résumé

The commercialization of generative AI is facing a critical inflection point as a potential price war looms. According to The Wall Street Journal, OpenAI is considering a significant cut to its token fees to compete with rival Anthropic, signaling a shift from a growth-at-all-costs model focused on token consumption. This move comes as both companies, reportedly losing billions on compute, prepare for IPOs, and as enterprise customers face "bill shock" from switching to usage-based token billing. Reports indicate poor ROI, with one analysis finding only 18 cents of every dollar spent on AI tokens generates user-facing value. The industry's initial phases—from flat-rate subscriptions to aggressive subsidies—have given way to a reckoning with real costs. Analysts debate the future: some predict a bifurcation between premium, high-cost models for complex tasks and cheaper alternatives for routine work, while others believe overall spending will still rise as agentic AI increases tokens per task. Notably, Chinese model DeepSeek's low-cost API is gaining traction with U.S. enterprises, adding competitive pressure. The core challenge is redefining value beyond token volume ("tokenmaxxing") toward measurable productivity ("valuemaxxing"), as the entire AI value chain, from cloud providers to chipmakers, feels the ripple effects of unsustainable pricing.

When a real Token price war erupts, how will the AI industry make money? The entire valuation logic for AI commercialization is at a moment where it needs to be rewritten. The era of competing on "cost-effectiveness" and "scarcity" may have arrived. For OpenAI, "the situation is deteriorating further." Analysis points out that "if OpenAI goes into decline, it could very likely drag down Nvidia, Oracle, Coreweave, and others."

The commercialization narrative of generative AI is facing its most profound self-examination in three years. From exchanging subsidies for users, to monthly subscription plans that hide costs, to Token-based billing exposing enterprise bill crises, the AI industry has completed a three-stage leap in commercialization within three years—and a potential price war could reset this entire monetization logic back to zero.

According to The Wall Street Journal, OpenAI is considering significantly lowering the Token fees it charges users, in order to compete for enterprise clients from rival Anthropic. Sources familiar with the matter stated that this move is partly to "seize the initiative," as OpenAI expects Anthropic to take similar price-cutting actions. OpenAI CEO Sam Altman recently admitted at an event that AI usage costs have become "a huge problem," and said the company will "help people get more value for less money."

The timing of this news is particularly sensitive. OpenAI secretly filed for an IPO this week, while Anthropic is also in the final countdown to its own listing. Meanwhile, Bloomberg's Silicon Data LLM Token Expenditure Index has fallen for seven consecutive trading days, marking its longest losing streak since January this year, reflecting the market's deep-seated anxiety about the sustainability of AI bills. The report bluntly states that a price war would directly erode the profit margins of both companies—and both are already losing tens of billions of dollars due to the massive computing power required by AI systems.

The core of this discussion is no longer just a single price-cut decision, but a more fundamental question: As the narrative of "more Token consumption is better" reaches its end, who will tell the next commercialization story for the AI industry, and how will it be told?

01

The Initial Three Stages: From Monthly Subsidies to Token Bills

The commercialization of generative AI has undergone a clear three-phase evolution in just three years.

Stage One: Monthly and Annual Subscriptions set the industry baseline. In February 2023, OpenAI launched ChatGPT Plus with a monthly fee of $19.99, pioneering paid C-end access for large models; Baidu, Alibaba, and Tencent followed suit, making fixed-fee subscriptions the standard for early-stage business models.

Stage Two: The subsidy war erupted in full force. To boost ARR (Annual Recurring Revenue), the core anchor for financing valuations, major players turned to large-scale subsidies: Google offered students 15 months of Gemini Advanced for free, OpenAI launched a Team membership plan at $1 for the first month, ByteDance's Doubao entered the market with pricing "99.3% lower than the industry standard," and Baidu announced its core models would be free. The essence of subsidies was trading losses for growth—reportedly, Microsoft lost an average of over $20 per user per month under the GitHub Copilot subscription model, with some heavy users costing up to $80 a month in losses.

Stage Three: The forced switch to usage-based billing. On June 1, 2026, Microsoft announced that all GitHub Copilot plans would officially transition to Token-based billing, converting the $19 monthly fee directly into an equivalent Token credit. This change brought the true costs long hidden by subscription models into the open—according to user calculations in the Reddit community, a single agent programming session could consume $30 to $40, depleting a monthly plan in just one use.

02

Bill Out of Control: When Tokens Cost More Than People

The implementation of Token-based pay-as-you-go billing fully revealed the true face of enterprise AI expenditure.

The numbers on the enterprise side are staggering. Uber COO Andrew Macdonald publicly stated in May 2026 that the link between the growth in Token consumption and substantive product improvement "does not yet exist," and even coined a term for it: "tokenmaxxing," describing employees performing valueless tasks to boost usage metrics.

More direct data: Uber exhausted its annual Token budget in just the first four months of 2026; Salesforce expects to pay Anthropic around $300 million for the full year.

Anthropic's own developer documentation shows that developers using Claude Code incur an average cost of about $13 per workday, with 90% of users having daily costs below $30—translating this, a 10-person development team could spend over $75,600 per year on Token fees alone.

The return on investment is equally alarming. Enterprise data platform Entelligence.AI, after aggregating data from 2,444 companies, found that for every $1 spent on AI Token fees, only 18 cents generated actual value that reached users; 44 cents were used to fix bugs introduced by the AI itself, 27 cents went towards rework, and 11 cents were consumed by review friction.

Faced with runaway bills, enterprises have begun proactive control measures. Amazon halted internal AI usage leaderboards, instructing employees "not to use AI just for the sake of using it"; Microsoft plans to gradually phase out Claude Code subscriptions for employees in some key product divisions. Goldman Sachs notes that some companies' spending on AI Tokens already accounts for 10% of their total employee labor costs, and this proportion could rise further in the coming quarters. This isn't about demand disappearing, but the end of the era of reckless AI spending.

03

Act Four: Price War Ignited, OpenAI Considers Major Price Cuts

It is against this backdrop that the fuse for a price war was lit.

According to The Wall Street Journal, Altman's consideration of price cuts was directly triggered by the pressure to catch up with Anthropic. Anthropic's revenue has grown significantly recently, its programming tool Claude Code has become popular among software engineers, and this five-year-old startup's valuation has even surpassed OpenAI's for the first time.

However, the cost of this price war will be exceptionally heavy. If prices are significantly lowered, it will further compress the already negative profit margins of both companies, and the competitive landscape offers very limited room for maneuver.

And a fundamental risk long identified by investors is that OpenAI and Anthropic's products are highly substitutable; clients can easily switch from one to the other—meaning price cuts, even if they retain customers in the short term, cannot truly build a moat and only delay market share loss.

This dilemma also transmits outwards through the financial cycle between cloud computing giants and AI labs.

According to corporate disclosure documents compiled by The Information, OpenAI and Anthropic together account for over half of the approximately $2 trillion in future cloud service commitments from Microsoft, Oracle, Google, and Amazon. If price cuts lead to downward revisions in revenue expectations, this transmission chain will face pressure from both ends.

American neuroscientist and AI expert Gary Marcus said: "This further exposes the fragility of OpenAI and indicates how serious the predicament it faces is. If OpenAI goes into decline, it could very likely drag down companies like Nvidia, Oracle, Coreweave. The situation is deteriorating rapidly."

A divergence of views is playing out openly on Wall Street. JPMorgan TMT analyst Mark Schilsky believes the current billing anxiety is merely a "minimum speed bump on the road to higher spending": if the average price per million tokens falls, but the penetration rate of paid AI among US companies continues to rise, overall token usage will mathematically increase substantially; coupled with agentic AI pushing single-task token consumption to several times that of traditional Q&A modes, long-term total expenditure is expected to be significantly higher than current levels.

Goldman Sachs semiconductor analyst Jim Covello holds a more pessimistic view, believing that the current industry chain prosperity has directed almost all value towards semiconductor companies, a phenomenon "unprecedented in history and unsustainable." Once enterprises face the true price of usage-based billing, the capital flows supporting GPU procurement and model training could face a reversal.

04

Act Five: The Next Story for Token Economics?

After the price war, the next chapter of AI industry commercialization has yet to be written, but its outline is emerging.

A report from Citadel Securities offers a directional framework: tiered pricing and charging based on scarcity. Its core logic is that inference-intensive frontier AI won't disappear but will increasingly concentrate in the hands of a few large enterprises capable of bearing the compute costs; for the broader range of enterprises, simpler models might be a more productive path until physical constraints ease. This implies AI usage will move towards stratification—high-value, complex tasks will continue to use frontier models, while routine and batch tasks will shift towards cheaper or local models.

JPMorgan holds a relatively optimistic judgment: even if the per-unit token price falls, the proliferation of agentic AI will multiply token consumption per task—existing data shows that after agentification, token consumption per task can become 3.5 times the original—potentially still expanding the overall expenditure scale. The current billing anxiety might just be a "minimum speed bump on the road to higher spending."

Nebius Chief Revenue Officer Marc Boroditsky proposed the concept of "valuemaxxing," advocating for the industry to shift from pursuing Token consumption maximization to making every Token truly generate value. This direction is gradually becoming an industry consensus—but true commercial implementation still requires AI labs to find a pricing system that both reflects the true cost and is acceptable to enterprise clients. This is precisely the core unresolved proposition in all current debates.

However, perhaps the most overlooked variable in this price war is Chinese models.

According to June data from American corporate spend management platform Ramp, DeepSeek has topped the list for growth in US enterprise software subscriptions. Ramp Chief Economist Ara Kharazian specifically emphasized that this is not about local deployment of open-source models, but rather "enterprises are directly sending and receiving data through DeepSeek," indicating real paid direct usage—he admitted "did not anticipate US companies would go and use DeepSeek." According to third-party estimates, DeepSeek V4-Pro's API price is about one-tenth that of GPT-5.5, and about one-eleventh that of Claude Opus 4.7.

As OpenAI and Anthropic, two tigers, fight, the ultimate beneficiary might be the player that has long embedded "accessible pricing" into its DNA and doesn't need to answer to IPO investors about profit margins. This may not be the most popular ending for this price war, but it is becoming an increasingly hard-to-ignore reality.

This article is from the WeChat public account "Hard AI," author: Xu Chao

Questions liées

QAccording to the article, what are the three main stages of generative AI's commercialization evolution?

AThe article outlines three stages: 1) Monthly/annual subscription plans establishing the baseline, like ChatGPT Plus. 2) A comprehensive subsidy war to drive user growth and valuations, with companies offering free or heavily discounted access. 3) The mandatory switch to usage-based (Token) billing, which exposed the true costs previously hidden by flat-rate subscriptions.

QWhat is 'tokenmaxxing' as mentioned in the article, and what problem does it highlight for businesses?

A'Tokenmaxxing' is a term coined by Uber's COO, describing the behavior where employees perform valueless tasks just to increase AI Token usage. It highlights the problem that increased Token consumption does not necessarily correlate with meaningful productivity gains or product improvements, leading to wasteful and uncontrolled AI spending for enterprises.

QWhy is a potential price war between OpenAI and Anthropic described as particularly costly and risky in the article?

AA price war is risky because both companies are already losing billions on the massive compute power required for AI systems. Lowering prices would further squeeze their negative profit margins. Furthermore, their products are highly substitutable, meaning price cuts may temporarily retain customers but do not build a sustainable competitive moat, potentially just delaying market share loss without solving the core profitability issue.

QWhat new pricing or usage frameworks are suggested as potential next chapters for AI commercialization after a price war?

AThe article suggests frameworks like tiered pricing based on scarcity/value, where advanced, compute-intensive models are reserved for high-value tasks at large enterprises, while simpler, cheaper models handle routine tasks. Another concept is 'valuemaxxing'—shifting the industry focus from maximizing Token consumption to ensuring each Token delivers actual value. Morgan Stanley also suggests that the rise of agentic AI could multiply Token use per task, potentially offsetting unit price declines.

QHow does the article characterize the role of Chinese AI models, specifically DeepSeek, in the current competitive landscape?

AThe article presents Chinese models like DeepSeek as a significant and often overlooked variable. It notes DeepSeek's API is priced at roughly one-tenth of GPT-5.5 and one-eleventh of Claude Opus. Data shows DeepSeek topping U.S. enterprise software subscription growth, indicating real paid adoption. The article suggests that while OpenAI and Anthropic compete, the ultimate beneficiary might be a player like DeepSeek, which has 'inclusive pricing' in its DNA and less immediate pressure to show profitability to IPO investors.

Lectures associées

Début de Warsh : le président de la Fed le plus au fait du Crypto de l'histoire apportera-t-il des surprises ou des chocs au marché ?

**Résumé :** Kevin Warsh, nouveau président de la Réserve fédérale américaine, s'apprête à tenir sa première conférence de presse monétaire. Sa nomination est historique : il est le premier président de la Fed à détenir personnellement des actifs numériques (investissements indirects dans Solana, dYdX, etc.), montrant une compréhension unique du secteur. Son dilemme est majeur : il doit faire face à une résurgence de l'inflation, qui exige une politique monétaire stricte (position "de faucon"), tout en répondant aux pressions politiques pour des baisses de taux. Parallèlement, son attitude envers les crypto-actifs diffère fondamentalement de celle de son prédécesseur. Il ne les considère pas comme de simples actifs spéculatifs, mais plutôt comme un "bon policier" pour la politique économique et une composante de la compétitivité américaine. Son impact potentiel sur le marché crypto s'articule autour de trois axes : 1. Un changement de paradigme réglementaire, passant de la prévention à l'intégration et à l'innovation. 2. Une reprixation des actifs liée aux taux d'intérêt, où sa clarté de communication pourrait réduire la prime d'incertitude. 3. Une légitimation accrue pouvant attirer les capitaux institutionnels traditionnels. Deux scénarios principaux sont envisagés pour sa première intervention : * **Scénario "Surprise"** : Un ton modéré ("de colombe") sur les taux combiné à des signaux favorables à l'innovation numérique pourrait booster le marché. * **Scénario "Choc"** : Un message excessivement restrictif sur les taux pourrait entraîner une vente généralisée des actifs risqués, y compris les cryptos. Bien qu'il ait dû vendre ses actifs crypto pour des raisons d'éthique, la compréhension intrinsèque de Warsh pour la technologie blockchain pourrait, à long terme, poser les bases d'une intégration plus structurelle des actifs numériques dans le système financier.

marsbitIl y a 3 h

Début de Warsh : le président de la Fed le plus au fait du Crypto de l'histoire apportera-t-il des surprises ou des chocs au marché ?

marsbitIl y a 3 h

XRP Ledger Lance le Rebranding XRPld Avec la Mise à Niveau Version 3.2.0

La version 3.2.0 du XRP Ledger (XRPL) est désormais disponible, introduisant une refonte majeure incluant le changement de nom du logiciel principal de « rippled » à « xrpld ». Cette mise à niveau se concentre principalement sur les améliorations des performances, de la sécurité et de l'évolutivité de l'infrastructure sous-jacente, plutôt que sur de nouvelles fonctionnalités utilisateur. Les principales avancées incluent des optimisations de mémoire pouvant réduire jusqu'à 40% l'utilisation de la mémoire serveur. Sur le plan de la sécurité, la modification `fixCleanup3_2_0` renforce plusieurs modules, notamment les coffres-forts à actif unique, le protocole de prêt, les échanges décentralisés et les jetons multi-usages. De nouveaux contrôles d'invariance garantissent la cohérence du registre après la suppression de comptes. Pour les développeurs, la mise à jour permet désormais de récupérer des informations sur les définitions du protocole et du serveur XRPL sans nécessiter de connexion active, facilitant ainsi la création de portefeuilles, d'explorateurs de blockchain et d'APIs. En termes d'évolutivité et de stabilité, les améliorations comprennent des tailles de bloc configurables, un stockage de base de données optimisé via nuDB, et le support optionnel de TLS/mutual TLS pour le serveur gRPC. Le port de peering par défaut est également passé du 51235 au 2459. Divers correctifs ont été apportés aux fonctions liées aux Market Makers Automatisés, aux paiements, aux séquestres de jetons et aux carnets d'ordres. Une note importante : les invariants de transaction ont été temporairement désactivés dans la v3.2.0 en raison d'un impact sur les performances, mais cela ne présente pas de risque pour la sécurité.

TheNewsCryptoIl y a 3 h

XRP Ledger Lance le Rebranding XRPld Avec la Mise à Niveau Version 3.2.0

TheNewsCryptoIl y a 3 h

L'AGI n'est pas l'arrivée, nouveau document de DeepMind : Vers l'ASI, le véritable progrès de l'IA ne fait que commencer

Si l'intelligence artificielle générale (IAG) était atteinte demain, quelle serait la prochaine étape ? Une étude de Google DeepMind suggère que l'IAG n'est pas un point final, mais une étape vers une superintelligence artificielle (ISA) dépassant les collectifs d'experts humains. L'étude distingue trois concepts : l'IAG (niveau médian humain), l'ISA (supérieure aux meilleurs collectifs humains dans presque tous les domaines) et l'IA universelle (limite théorique). Elle propose quatre voies potentielles vers l'ISA : 1. **Extension des ressources** : augmentation de la puissance de calcul, des données et des modèles. 2. **Évolution algorithmique** : améliorations incrémentales ou nouveaux paradigmes (apprentissage continu, utilisation d'outils, modèles du monde). 3. **Auto-amélioration récursive** : des IA plus performantes conçoivent la génération suivante, créant une boucle de rétroaction positive. 4. **Coordination multi-agents** : des systèmes IAG collaborant atteignent une intelligence collective supérieure. L'étude identifie six principaux goulets d'étranglement : 1. **Le mur des données** : les données humaines de haute qualité pourraient s'épuiser. 2. **Pressions économiques et ressources naturelles** : coûts énergétiques et matériels. 3. **Limites des paradigmes neuronaux actuels** : problèmes d'apprentissage continu, de raisonnement robuste, d'hallucinations. 4. **Difficulté croissante de la recherche**. 5. **Barrières à l'abstraction** : difficulté à former de nouveaux concepts fondamentaux. 6. **Régulation, gouvernance et réaction sociale**. Un défi crucial est l'évaluation des capacités de l'IA au-delà du niveau humain, nécessitant de nouveaux benchmarks. L'étude conclut que la progression vers l'ISA reste incertaine, soumise à des contraintes physiques et de ressources, et appelle à un effort de recherche interdisciplinaire pour mieux anticiper cette évolution.

marsbitIl y a 4 h

L'AGI n'est pas l'arrivée, nouveau document de DeepMind : Vers l'ASI, le véritable progrès de l'IA ne fait que commencer

marsbitIl y a 4 h

Trading

Spot
Futures

Articles tendance

Comment acheter WAR

Bienvenue sur HTX.com ! Nous vous permettons d'acheter WAR (WAR) de manière simple et pratique. Suivez notre guide étape par étape pour commencer votre parcours crypto.Étape 1 : Création de votre compte HTXUtilisez votre adresse e-mail ou votre numéro de téléphone pour ouvrir un compte sur HTX gratuitement. L'inscription se fait en toute simplicité et débloque toutes les fonctionnalités.Créer mon compteÉtape 2 : Choix du mode de paiement (rubrique Acheter des cryptosCarte de crédit/débit : utilisez votre carte Visa ou Mastercard pour acheter instantanément WAR (WAR).Solde :utilisez les fonds du solde de votre compte HTX pour trader en toute simplicité.Prestataire tiers :pour accroître la commodité d'utilisation, nous avons ajouté des modes de paiement populaires tels que Google Pay et Apple Pay.P2P :tradez directement avec d'autres utilisateurs sur HTX.OTC (de gré à gré) : nous offrons des services personnalisés et des taux de change compétitifs aux traders.Étape 3 : stockage de vos WAR (WAR)Après avoir acheté vos WAR (WAR), stockez-les sur votre compte HTX. Vous pouvez également les envoyer ailleurs via un transfert sur la blockchain ou les utiliser pour trader d'autres cryptos.Étape 4 : tradez des WAR (WAR)Tradez facilement WAR (WAR) sur le marché Spot de HTX. Il vous suffit d'accéder à votre compte, de sélectionner la paire de trading, d'exécuter vos trades et de les suivre en temps réel. Nous offrons une expérience conviviale aux débutants comme aux traders chevronnés.

532 vues totalesPublié le 2024.12.11Mis à jour le 2026.06.02

Comment acheter WAR

Discussions

Bienvenue dans la Communauté HTX. Ici, vous pouvez vous tenir informé(e) des derniers développements de la plateforme et accéder à des analyses de marché professionnelles. Les opinions des utilisateurs sur le prix de WAR (WAR) sont présentées ci-dessous.

活动图片