Canonical URL: https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/ --- description: Tether’s own guide shows light usage costing less through an API than a new laptop. Volume, capacity and quality determine when local AI is worthwhile. title: QVAC: AI without a per-token bill still has costs image: https://elacelerador.es/images/social/v1/qvac-ai-without-per-token-bill-still-costs-e242f88c5b09042528feaf275594ee38.jpg --- 1. [Home](https://elacelerador.es/en/) 2. [Cases and research](https://elacelerador.es/en/cases-and-research/) # QVAC: AI without a per-token bill still has costs Tether’s own guide shows light usage costing less through an API than a new laptop. Volume, capacity and quality determine when local AI is worthwhile. By [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 at 09:14 CEST 6 min read [ Add as preferred source ](https://www.google.com/preferences/source?q=elacelerador.es&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) Share [ ](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dlinkedin%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on LinkedIn") [ ](https://wa.me/?text=QVAC%3A+AI+without+a+per-token+bill+still+has+costs+https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dwhatsapp%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on WhatsApp") [ ](https://t.me/share/url?url=https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dtelegram%26utm%5Fmedium%3Dsocial&text=QVAC%3A+AI+without+a+per-token+bill+still+has+costs&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Telegram") [ ](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dfacebook%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Facebook") ![A blue laptop motherboard with chips, memory slots, a fan and a copper cooling pipe.](https://elacelerador.es/images/variants/v1/qvac-ia-factura-token-tambien-tiene-costes-placa-base-azul-portatil-chips-w1280.webp) A Sony VAIO laptop motherboard photographed by Hannes Grobe in November 2020\. Archival image providing context for processing on local devices. Photo: [Hannes Grobe / Wikimedia Commons, CC BY-SA 4.0](https://commons.wikimedia.org/wiki/File:Sony-vaio-mainboard%5Fhg.jpg?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) **In this report** 1. [A free SDK leaves purchasing decisions to resolve](#article-section-a-free-sdk-leaves-purchasing-decisions-to-resolve) 2. [The $7 versus $27 example has specific conditions](#article-section-the-7-versus-27-example-has-specific-conditions) 3. [Quality and capacity change the economics of the same volume](#article-section-quality-and-capacity-change-the-economics-of-the-same-volume) 2 figures * [Managing spending remains a cloud challenge](#article-figure-1) * [Reported barriers extend beyond processing prices](#article-figure-2) [Download data (CSV)](https://elacelerador.es/en/download.csv?article=%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F) Chart values, table cells, source references and the report revision date. QVAC’s own October 12 cost guide contains a comparison that limits the promise of local savings. For light chat use, it calculates about $7 a month through a mid-range API against $27 when amortizing a laptop bought for the work. Removing the per-token bill can change who carries the cost without making local processing the cheapest option. QVAC is Tether’s local AI SDK , available as open software for running models on devices. Its expansion depends on explaining which workloads benefit. A developer may value offline continuity or control over data even when the alternative costs more. The commercial argument becomes clearer when it separates those benefits from financial savings that require their own evidence. ## A free SDK leaves purchasing decisions to resolve The repository and its Apache 2.0 license make the code’s availability verifiable. That condition describes the SDK, rather than universal terms for every component connected to it. Models carry their own conditions of use. Installing free software also leaves the cost of hardware, available capacity, maintenance and the work of integrating a feature into an application. The data behind the story ### Managing spending remains a cloud challenge Three selected categories from the Flexera 2026 report. The source describes fieldwork as winter 2025; a worldwide panel of 753 technical professionals and executive leaders. Respondents reporting the challenge Tap or hover for details. All values are also available in the data table. Expand notesCollapse notes **Sources:** Statista; Flexera 2026 State of the Cloud Report, Figure 22 An independent panel of vetted respondents, not a census: 62% of participants from the US, 23% Europe and 13% Asia-Pacific. The source does not publish exact fieldwork days. Rounded percentages and non-exclusive categories. These are not observed savings or a measurement of QVAC or Spanish companies. Explore the data table Scroll left or right to see all columns __Managing spending remains a cloud challenge (%)__ | Category | Respondents reporting the challenge | | -------------------------- | ----------------------------------- | | Managing cloud spending | 85% | | Security | 82% | | Managing software licenses | 78% | The Flexera 2026 report places spending, security and licenses among the challenges reported by the people surveyed. It does not measure demand for local AI or savings from adopting it. It does help frame the buyer’s question: which cost is removed, which resource replaces it and which obligations remain. Moving from a consumption invoice to owned hardware changes the structure of spending. Exposure to idle capacity changes too. An API can fit a small, irregular workload. Purchased hardware retains its cost while processing few requests; hardware already available creates a different decision about spare capacity and alternative uses. The same workload can justify different choices depending on the starting point. ## The $7 versus $27 example has specific conditions QVAC’s published comparison uses 20 daily requests over 22 working days and about 2.8 million tokens a month. It assumes 94% input tokens and 6% output, at list prices of $2 and $10 per million respectively. Its example is a 16 GB MacBook Air costing $1,299, amortized over four years. These are the provider’s assumptions in US dollars, rather than a quote for Spain or an observed customer invoice. The calculation excludes electricity and is not a comparison of equivalent capabilities. The chosen local model, remote model and device can perform different work or need different corrections. List prices also exclude cache discounts. These conditions matter as much as the final number: a break-even point obtained with one token mix does not automatically apply to another. The guide itself shows the choice changing with volume and service type. In its heavy-chat scenario, with 100 daily requests and 14 million tokens a month, it attributes $35 to a mid-range API and $8 to Qwen3.8-27B served by Alibaba in the cloud, against the same $27 laptop. A useful comparison must retain the service’s identity and the task it can perform; combining every API into a single category conceals differences in price and quality. For Tether, that distinction supports a more precise commercial presentation than a general promise of free operation. QVAC can remove a remote provider’s inference charge from a particular path. The final advantage depends on how often that path is used, how long it takes and which result the process accepts. Local control can be a reason to buy in its own right, without being presented as financial savings. ## Quality and capacity change the economics of the same volume QVAC’s getting-started guide acknowledges memory limits and differences from the largest cloud models on difficult reasoning, broad knowledge and long contexts. It also explains that the first download requires an internet connection. These conditions help select a task and limit the usefulness of comparing only the number of tokens processed. An independent preprint on the Bench360 framework , published in November 2025, evaluates answer quality, latency and energy across local inference configurations. It uses four engines and three types of 24 GB GPU, under individual-request, batch and concurrent workloads. It finds that the suitable configuration depends on the task and load. It does not evaluate QVAC or establish its laptop performance; it contributes a comparison method that goes beyond speed or memory alone. For a business process, the economic unit can be an accepted result delivered within the required time. Total cost would include computing resources, integration and necessary review, divided by the tasks that meet the agreed criterion. This definition helps establish whether an option that is cheap at generating text remains cheap after correction. Latency also matters when the same device has other work to do. The data behind the story ### Reported barriers extend beyond processing prices Spain, July to September 2024\. Four barrier categories in Deloitte’s online survey of 100 technology and business leaders involved in piloting or implementing generative AI. Respondents Tap or hover for details. All values are also available in the data table. Expand notesCollapse notes **Sources:** Statista; Deloitte: original Spain findings; Deloitte: fourth-wave methodology, page 4 A selected sample of AI-experienced leaders, with multiple responses. It does not represent all Spanish companies or measure incidents or willingness to buy QVAC. Fieldwork in 2024, distinct from Flexera’s worldwide panel; the two charts are not a time series. Explore the data table Scroll left or right to see all columns __Reported barriers extend beyond processing prices (%)__ | Category | Respondents | | -------------------------- | ----------- | | Regulatory compliance | 46% | | Risk management | 36% | | Lack of a governance model | 31% | | Lack of talent and skills | 23% | Deloitte’s Spanish sample identifies compliance, risk, governance and skills barriers. It cannot quantify deployment costs, but it shows why the decision extends beyond the processing invoice. Keeping inference on a device can avoid sending it to a server; a connected tool can still transmit information. QVAC’s own cost guide distinguishes the model from the tools connected to it. A defensible adoption decision first specifies the task, available hardware and required result, then compares costs at a compatible volume. Each model’s prices and conditions, external connections and review requirements form part of that decision. The same discipline applies to local and remote options, without assuming which will win. QVAC’s opportunity is to turn technical autonomy into an understandable economic choice. An open SDK makes it easier to start; a verifiable account of capacity, quality and cost helps determine where to keep using it. The strongest argument for expansion will identify the tasks where that combination delivers value, even when the decisive benefit is offline operation rather than lower spending. [Back to top ↑](#article-top) ## In this article ### Sources and documents * [QVAC: SDK, local inference and supported model formats](https://qvac.tether.io/products/sdk?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Tether / QVAC * [QVAC: economics of token consumption, October 12, 2026](https://qvac.tether.io/blog/local-ai-cost-per-token?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Tether / QVAC * [QVAC: getting started and constraints, July 22, 2026](https://qvac.tether.io/blog/quick-start-build-your-first-local-ai-app-with-your-coding-agent?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Tether / QVAC * [QVAC: official source repository](https://github.com/tetherto/qvac?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Tether / GitHub * [QVAC SDK: Apache License 2.0](https://raw.githubusercontent.com/tetherto/qvac/refs/heads/main/LICENSE?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Tether Data * Statista * [Flexera 2026 State of the Cloud: Figure 22 and methodology](https://info.flexera.com/CM-REPORT-State-of-the-Cloud?lead%5Fsource=Organic+Search&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es#methodology) · Flexera * [Deloitte: original findings for Spain](https://www.deloitte.com/es/es/services/consulting/research/estado-ia-generativa-empresas-espana.html?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Deloitte * [Deloitte: fourth-wave field methodology, page 4](https://www.deloitte.com/content/dam/assets-zone4/latam/es/docs/services/consulting/2025/IA-Generativa-en-la-Empresa-de-la-expectativa-al-valor-sostenible-Nov25.pdf?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · Deloitte * [Stuhlmann et al.: Bench360 preprint, November 12, 2025](https://arxiv.org/html/2511.16682v1?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · arXiv / authors Share [ ](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dlinkedin%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on LinkedIn") [ ](https://wa.me/?text=QVAC%3A+AI+without+a+per-token+bill+still+has+costs+https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dwhatsapp%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on WhatsApp") [ ](https://t.me/share/url?url=https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dtelegram%26utm%5Fmedium%3Dsocial&text=QVAC%3A+AI+without+a+per-token+bill+still+has+costs&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Telegram") [ ](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Felacelerador.es%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F%3Futm%5Fsource%3Dfacebook%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Facebook") ![Portrait of Marian Vasile](https://elacelerador.es/images/variants/v1/marian-vasile-autor-acelerador-retrato-w320.webp) [LinkedIn](https://www.linkedin.com/in/nairamelisav/?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) About the author ## [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) Marian Vasile is a MarTech and growth professional. His work spans product marketing, SEO, paid advertising and deep online tracking, including behavioral analytics and attribution. ## Related articles [Cases and research](https://elacelerador.es/en/cases-and-research/) Analysis ### [iSpeedToLead: the next purchase needs clear refund terms](https://elacelerador.es/en/cases-and-research/ispeedtolead-next-purchase-clear-refund-terms/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 6 min [Cases and research](https://elacelerador.es/en/cases-and-research/) Analysis ### [Tucuvi: a successful call needs follow-up that reaches completion](https://elacelerador.es/en/cases-and-research/tucuvi-successful-call-patient-follow-up/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 5 min [Cases and research](https://elacelerador.es/en/cases-and-research/) Analysis ### [Social Discovery Group: charging for conversation requires clarity about the purchase](https://elacelerador.es/en/cases-and-research/social-discovery-group-conversations-trust-credits/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 5 min **In this report** 1. [A free SDK leaves purchasing decisions to resolve](#article-section-a-free-sdk-leaves-purchasing-decisions-to-resolve) 2. [The $7 versus $27 example has specific conditions](#article-section-the-7-versus-27-example-has-specific-conditions) 3. [Quality and capacity change the economics of the same volume](#article-section-quality-and-capacity-change-the-economics-of-the-same-volume) [Download data (CSV)](https://elacelerador.es/en/download.csv?article=%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F) [Back to top ↑](#article-top) ## Related articles [Cases and research](https://elacelerador.es/en/cases-and-research/) Analysis ### [iSpeedToLead: the next purchase needs clear refund terms](https://elacelerador.es/en/cases-and-research/ispeedtolead-next-purchase-clear-refund-terms/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 6 min [Cases and research](https://elacelerador.es/en/cases-and-research/) Analysis ### [Tucuvi: a successful call needs follow-up that reaches completion](https://elacelerador.es/en/cases-and-research/tucuvi-successful-call-patient-follow-up/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 5 min [Cases and research](https://elacelerador.es/en/cases-and-research/) Analysis ### [Social Discovery Group: charging for conversation requires clarity about the purchase](https://elacelerador.es/en/cases-and-research/social-discovery-group-conversations-trust-credits/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 19 October 2026 5 min ```json {"@context":"https://schema.org","@graph":[{"@type":["Organization","NewsMediaOrganization"],"@id":"https://elacelerador.es/#publisher","name":"El Acelerador","url":"https://elacelerador.es/","description":"Marketing and growth hacking with examples from Spanish companies.","email":"pr@elacelerador.es","publishingPrinciples":"https://elacelerador.es/en/editorial-policy/","correctionsPolicy":"https://elacelerador.es/en/editorial-policy/","verificationFactCheckingPolicy":"https://elacelerador.es/en/editorial-policy/","masthead":"https://elacelerador.es/en/about/","ownershipFundingInfo":"https://elacelerador.es/en/about/","logo":{"@type":"ImageObject","url":"https://elacelerador.es/icon-512.png?v=20261017-stacked"}},{"@type":"Person","@id":"https://elacelerador.es/author/marian-vasile/#person","name":"Marian Vasile","url":"https://elacelerador.es/en/author/marian-vasile/","description":"Marian Vasile is a MarTech and growth professional. His work spans product marketing, SEO, paid advertising and deep online tracking, including behavioral analytics and attribution.","sameAs":["https://marianvasile.com/","https://www.linkedin.com/in/nairamelisav/"],"image":{"@type":"ImageObject","url":"https://elacelerador.es/images/marian-vasile-autor-acelerador-retrato.webp","contentUrl":"https://elacelerador.es/images/marian-vasile-autor-acelerador-retrato.webp","width":519,"height":519,"caption":"Marian Vasile"}},{"@type":"ImageObject","url":"https://elacelerador.es/images/qvac-ia-factura-token-tambien-tiene-costes-placa-base-azul-portatil-chips.webp","contentUrl":"https://elacelerador.es/images/qvac-ia-factura-token-tambien-tiene-costes-placa-base-azul-portatil-chips.webp","width":1920,"height":1592,"caption":"A Sony VAIO laptop motherboard photographed by Hannes Grobe in November 2020. Archival image providing context for processing on local devices.","@id":"https://elacelerador.es/images/qvac-ia-factura-token-tambien-tiene-costes-placa-base-azul-portatil-chips.webp#image","creditText":"Hannes Grobe / Wikimedia Commons, CC BY-SA 4.0","isBasedOn":"https://commons.wikimedia.org/wiki/File:Sony-vaio-mainboard_hg.jpg"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-1","name":"QVAC: SDK, local inference and supported model formats","url":"https://qvac.tether.io/products/sdk","creditText":"Tether / QVAC"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-2","name":"QVAC: economics of token consumption, October 12, 2026","url":"https://qvac.tether.io/blog/local-ai-cost-per-token","creditText":"Tether / QVAC"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-3","name":"QVAC: getting started and constraints, July 22, 2026","url":"https://qvac.tether.io/blog/quick-start-build-your-first-local-ai-app-with-your-coding-agent","creditText":"Tether / QVAC"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-4","name":"QVAC: official source repository","url":"https://github.com/tetherto/qvac","creditText":"Tether / GitHub"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-5","name":"QVAC SDK: Apache License 2.0","url":"https://raw.githubusercontent.com/tetherto/qvac/refs/heads/main/LICENSE","creditText":"Tether Data"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-6","name":"Statista","url":"https://www.statista.com/"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-7","name":"Flexera 2026 State of the Cloud: Figure 22 and methodology","url":"https://info.flexera.com/CM-REPORT-State-of-the-Cloud?lead_source=Organic%20Search#methodology","creditText":"Flexera"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-8","name":"Deloitte: original findings for Spain","url":"https://www.deloitte.com/es/es/services/consulting/research/estado-ia-generativa-empresas-espana.html","creditText":"Deloitte"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-9","name":"Deloitte: fourth-wave field methodology, page 4","url":"https://www.deloitte.com/content/dam/assets-zone4/latam/es/docs/services/consulting/2025/IA-Generativa-en-la-Empresa-de-la-expectativa-al-valor-sostenible-Nov25.pdf","creditText":"Deloitte"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-10","name":"Stuhlmann et al.: Bench360 preprint, November 12, 2025","url":"https://arxiv.org/html/2511.16682v1","creditText":"arXiv / authors"},{"@type":"Dataset","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#dataset","url":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-top","name":"The data behind the story: QVAC: AI without a per-token bill still has costs","description":"Data accompanying the article, with the published research figures, tables, sources and methodological notes. The CSV download contains the complete article data export.","inLanguage":"en-GB","isPartOf":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article","hasPart":[{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-figure-1-dataset"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-figure-2-dataset"}],"isAccessibleForFree":false,"isBasedOn":["https://www.statista.com/","https://info.flexera.com/CM-REPORT-State-of-the-Cloud?lead_source=Organic%20Search#methodology","https://www.deloitte.com/es/es/services/consulting/research/estado-ia-generativa-empresas-espana.html","https://www.deloitte.com/content/dam/assets-zone4/latam/es/docs/services/consulting/2025/IA-Generativa-en-la-Empresa-de-la-expectativa-al-valor-sostenible-Nov25.pdf"],"publisher":{"@id":"https://elacelerador.es/#publisher"},"distribution":{"@type":"DataDownload","name":"Download data (CSV)","contentUrl":"https://elacelerador.es/en/download.csv?article=%2Fen%2Fcases-and-research%2Fqvac-local-ai-token-cost-amortization%2F","encodingFormat":"text/csv","requiresSubscription":true,"isAccessibleForFree":false}},{"@type":"Dataset","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-figure-1-dataset","url":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-figure-1","name":"Managing spending remains a cloud challenge","description":"Three selected categories from the Flexera 2026 report. The source describes fieldwork as winter 2025; a worldwide panel of 753 technical professionals and executive leaders. An independent panel of vetted respondents, not a census: 62% of participants from the US, 23% Europe and 13% Asia-Pacific. The source does not publish exact fieldwork days. Rounded percentages and non-exclusive categories. These are not observed savings or a measurement of QVAC or Spanish companies. Flexera 2026 State of the Cloud, Figure 22: 753 technical professionals and executive leaders from an independent vetted panel. Source identifies fieldwork as winter 2025 without exact dates. ISO coverage represents bounds of the reported year, not known interview days. Three selected categories, percentages rounded to the nearest whole number. Statista title uses 2025; the consulted original is the 2026 edition.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#dataset"},"isAccessibleForFree":false,"publisher":{"@id":"https://elacelerador.es/#publisher"},"variableMeasured":[{"@type":"PropertyValue","name":"Respondents reporting the challenge","unitText":"%"}],"isBasedOn":["https://www.statista.com/","https://info.flexera.com/CM-REPORT-State-of-the-Cloud?lead_source=Organic%20Search#methodology"],"provider":{"@type":"Organization","name":"Flexera","url":"https://www.flexera.com/"},"temporalCoverage":"2025-01-01/2025-12-31","spatialCoverage":{"@type":"Place","name":"Worldwide","identifier":"GLOBAL"},"measurementTechnique":"Flexera 2026 State of the Cloud, Figure 22: 753 technical professionals and executive leaders from an independent vetted panel. Source identifies fieldwork as winter 2025 without exact dates. ISO coverage represents bounds of the reported year, not known interview days. Three selected categories, percentages rounded to the nearest whole number. Statista title uses 2025; the consulted original is the 2026 edition."},{"@type":"Dataset","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-figure-2-dataset","url":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-figure-2","name":"Reported barriers extend beyond processing prices","description":"Spain, July to September 2024. Four barrier categories in Deloitte’s online survey of 100 technology and business leaders involved in piloting or implementing generative AI. A selected sample of AI-experienced leaders, with multiple responses. It does not represent all Spanish companies or measure incidents or willingness to buy QVAC. Fieldwork in 2024, distinct from Flexera’s worldwide panel; the two charts are not a time series. Online survey of 100 Spain technology and business leaders experienced in AI and directly involved in piloting or implementing generative AI. Fourth wave, July-September 2024; ISO bounds express those months, not exact interview dates. Global wave: 2,773 people in 14 countries and six industries. Selected categories and multiple responses.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#dataset"},"isAccessibleForFree":false,"publisher":{"@id":"https://elacelerador.es/#publisher"},"variableMeasured":[{"@type":"PropertyValue","name":"Respondents","unitText":"%"}],"isBasedOn":["https://www.statista.com/","https://www.deloitte.com/es/es/services/consulting/research/estado-ia-generativa-empresas-espana.html","https://www.deloitte.com/content/dam/assets-zone4/latam/es/docs/services/consulting/2025/IA-Generativa-en-la-Empresa-de-la-expectativa-al-valor-sostenible-Nov25.pdf"],"provider":{"@type":"Organization","name":"Deloitte","url":"https://www.deloitte.com/"},"temporalCoverage":"2024-07-01/2024-09-30","spatialCoverage":{"@type":"Place","name":"Spain","identifier":"ES"},"measurementTechnique":"Online survey of 100 Spain technology and business leaders experienced in AI and directly involved in piloting or implementing generative AI. Fourth wave, July-September 2024; ISO bounds express those months, not exact interview dates. Global wave: 2,773 people in 14 countries and six industries. Selected categories and multiple responses."},{"@type":"Article","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article","url":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/","headline":"QVAC: AI without a per-token bill still has costs","description":"Tether’s own guide shows light usage costing less through an API than a new laptop. Volume, capacity and quality determine when local AI is worthwhile.","datePublished":"2026-10-19T07:14:49.666Z","dateModified":"2026-10-19T07:14:49.666Z","inLanguage":"en-GB","mainEntityOfPage":{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#webpage"},"translationOfWork":{"@id":"https://elacelerador.es/casos-y-estudios/qvac-ia-local-coste-token-amortizacion/#article"},"isPartOf":{"@id":"https://elacelerador.es/en/#website"},"isAccessibleForFree":false,"hasPart":[{"@type":"WebPageElement","isAccessibleForFree":false,"cssSelector":".reader-restricted"}],"image":[{"@id":"https://elacelerador.es/images/qvac-ia-factura-token-tambien-tiene-costes-placa-base-azul-portatil-chips.webp#image"}],"articleSection":"Cases and research","wordCount":1144,"timeRequired":"PT6M","citation":[{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-1"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-2"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-3"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-4"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-5"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-6"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-7"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-8"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-9"},{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article-source-10"}],"author":[{"@id":"https://elacelerador.es/author/marian-vasile/#person"}],"publisher":{"@id":"https://elacelerador.es/#publisher"}},{"@type":"WebPage","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#webpage","url":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/","name":"QVAC: AI without a per-token bill still has costs","description":"Tether’s own guide shows light usage costing less through an API than a new laptop. Volume, capacity and quality determine when local AI is worthwhile.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/#website"},"mainEntity":{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#article"},"breadcrumb":{"@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#breadcrumbs"},"primaryImageOfPage":{"@id":"https://elacelerador.es/images/qvac-ia-factura-token-tambien-tiene-costes-placa-base-azul-portatil-chips.webp#image"}},{"@type":"BreadcrumbList","@id":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/#breadcrumbs","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://elacelerador.es/en/"},{"@type":"ListItem","position":2,"name":"Cases and research","item":"https://elacelerador.es/en/cases-and-research/"},{"@type":"ListItem","position":3,"name":"QVAC: AI without a per-token bill still has costs","item":"https://elacelerador.es/en/cases-and-research/qvac-local-ai-token-cost-amortization/"}]}]} ```