{"id":427,"date":"2026-09-01T10:49:40","date_gmt":"2026-09-01T10:49:40","guid":{"rendered":"https:\/\/wrangleai.com\/blog\/?p=427"},"modified":"2026-09-01T10:49:42","modified_gmt":"2026-09-01T10:49:42","slug":"how-to-track-ai-usage-across-openai-anthropic-and-gemini","status":"publish","type":"post","link":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/","title":{"rendered":"How to Track AI Usage Across OpenAI, Anthropic and Gemini"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/wrangleai.com\/blog\/why-enterprises-are-struggling-to-track-ai-usage\/\">Enterprise AI<\/a> infrastructure is changing quickly. Many organisations no longer rely on a single model provider. One engineering team may use OpenAI for a customer-facing assistant, another may use <a href=\"https:\/\/www.anthropic.com\/\">Anthropic Claude<\/a> for coding or complex reasoning, while another product uses <a href=\"https:\/\/gemini.google.com\/\">Google Gemini<\/a> for multimodal workloads.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This multi-provider approach gives teams more choice, but it creates a new operational problem: <strong><a href=\"https:\/\/wrangleai.com\/blog\/why-enterprises-are-struggling-to-track-ai-usage\/\">How do you track AI usage across all of these providers?<\/a><\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Each provider has its own console, projects, API keys, usage reports, billing structure, token measurements and monitoring tools. OpenAI provides API usage reporting at organisation and project level. Anthropic provides usage and cost reporting across workspaces, models and API keys. Google provides Gemini API usage information through Google AI Studio and its wider billing environment.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">These tools are useful when teams operate inside one ecosystem. The challenge begins when an organisation needs one clear view across all of them.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/wrangleai.com\/\">WrangleAI<\/a> helps organisations centralise visibility across AI providers, models, keys, applications and agents. Instead of treating AI usage as separate provider reports, teams can build a more complete operational view of how AI is being consumed across the business.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This guide explains how to track AI usage across OpenAI, Anthropic and Gemini, which metrics matter, where provider dashboards fall short, and how organisations can build a stronger multi-provider monitoring strategy.<\/p>\n\n\n<div class=\"wp-block-aioseo-table-of-contents\"><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-why-ai-usage-tracking-has-become-more-important\">Why AI Usage Tracking Has Become More Important<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-costs-become-distributed\">Costs Become Distributed<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-model-usage-becomes-harder-to-understand\">Model Usage Becomes Harder to Understand<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-ownership-becomes-unclear\">Ownership Becomes Unclear<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-ai-agents-can-increase-consumption-rapidly\">AI Agents Can Increase Consumption Rapidly<\/a><\/li><\/ul><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-what-does-ai-usage-actually-mean\">What Does AI Usage Actually Mean?<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-token-usage\">Token Usage<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-request-volume\">Request Volume<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-cost\">Cost<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-model-usage\">Model Usage<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-application-and-agent-usage\">Application and Agent Usage<\/a><\/li><\/ul><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-to-track-openai-api-usage\">How to Track OpenAI API Usage<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-use-the-openai-usage-dashboard\">Use the OpenAI Usage Dashboard<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-track-tokens-from-api-responses\">Track Tokens from API Responses<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-export-openai-usage-data\">Export OpenAI Usage Data<\/a><\/li><\/ul><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-to-track-anthropic-claude-usage\">How to Track Anthropic Claude Usage<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-monitor-usage-by-model\">Monitor Usage by Model<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-monitor-usage-by-api-key\">Monitor Usage by API Key<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-track-input-and-output-tokens\">Track Input and Output Tokens<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-export-anthropic-usage\">Export Anthropic Usage<\/a><\/li><\/ul><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-to-track-google-gemini-usage\">How to Track Google Gemini Usage<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-track-gemini-usage-through-ai-studio\">Track Gemini Usage Through AI Studio<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-monitor-gemini-costs\">Monitor Gemini Costs<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-use-gemini-api-logs-where-appropriate\">Use Gemini API Logs Where Appropriate<\/a><\/li><\/ul><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-the-problem-with-tracking-each-provider-separately\">The Problem with Tracking Each Provider Separately<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-create-a-common-ai-usage-data-model\">Create a Common AI Usage Data Model<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-provider\">Provider<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-model\">Model<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-application\">Application<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-team\">Team<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-api-key-or-identity\">API Key or Identity<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-agent\">Agent<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-cost-2\">Cost<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-time\">Time<\/a><\/li><\/ul><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-use-separate-keys-and-identities-for-better-attribution\">Use Separate Keys and Identities for Better Attribution<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-monitor-ai-usage-by-application-not-just-provider\">Monitor AI Usage by Application, Not Just Provider<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-track-ai-usage-by-agent\">Track AI Usage by Agent<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-monitor-ai-costs-across-providers\">Monitor AI Costs Across Providers<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-build-alerts-for-unexpected-ai-usage\">Build Alerts for Unexpected AI Usage<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-track-which-model-actually-processes-each-request\">Track Which Model Actually Processes Each Request<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-connect-ai-usage-monitoring-with-governance\">Connect AI Usage Monitoring with Governance<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-create-a-central-ai-usage-dashboard\">Create a Central AI Usage Dashboard<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-wrangleai-helps-track-ai-usage-across-providers\">How WrangleAI Helps Track AI Usage Across Providers<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-final-thoughts\">Final Thoughts<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-faqs\">FAQs<\/a><ul><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\">How can I track AI usage across OpenAI, Anthropic and Gemini?<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\">What AI usage metrics should engineering teams track?<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\">Why are provider dashboards not enough for multi-provider AI?<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\">How does WrangleAI help track AI usage?<\/a><\/li><li><a class=\"aioseo-toc-item\" href=\"#aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\">Can AI usage monitoring help reduce AI costs?<\/a><\/li><\/ul><\/li><\/ul><\/div>\n\n\n<h2 id=\"aioseo-why-ai-usage-tracking-has-become-more-important\" class=\"wp-block-heading\">Why AI Usage Tracking Has Become More Important<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">AI started in many organisations as a small experiment. A development team received an API key, connected a model and tested a few use cases.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That environment can become complicated very quickly. The same company may soon have customer support assistants, internal search tools, AI coding workflows, document analysis systems, marketing tools, data applications and autonomous agents. Each workload may use a different model.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This creates several problems for engineering, finance and governance teams.<\/p>\n\n\n\n<h3 id=\"aioseo-costs-become-distributed\" class=\"wp-block-heading\">Costs Become Distributed<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">AI spending may exist across several providers and projects.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Finance may know the total amount being paid to each provider, but engineering teams need more detail. They need to know which product, application, model or agent is responsible for that cost.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Without workload-level visibility, cost optimisation becomes difficult.<\/p>\n\n\n\n<h3 id=\"aioseo-model-usage-becomes-harder-to-understand\" class=\"wp-block-heading\">Model Usage Becomes Harder to Understand<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Teams may have approved certain models for production while developers experiment with others.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Without central monitoring, it becomes harder to understand which models are actually being used across the organisation.<\/p>\n\n\n\n<h3 id=\"aioseo-ownership-becomes-unclear\" class=\"wp-block-heading\">Ownership Becomes Unclear<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">An unexpected increase in AI usage means very little if nobody knows which application generated it.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Effective monitoring needs to connect consumption with ownership.<\/p>\n\n\n\n<h3 id=\"aioseo-ai-agents-can-increase-consumption-rapidly\" class=\"wp-block-heading\">AI Agents Can Increase Consumption Rapidly<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Agentic systems make this challenge more serious because one user action can trigger several model calls.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">An agent may analyse a task, call a tool, review the result, call another model and continue the workflow until it reaches its goal.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A small increase in agent activity can therefore create a much larger increase in model usage.<\/p>\n\n\n\n<h2 id=\"aioseo-what-does-ai-usage-actually-mean\" class=\"wp-block-heading\">What Does AI Usage Actually Mean?<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Before monitoring AI usage, organisations need to define what they want to measure.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Usage should not be reduced to one number.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A useful monitoring programme normally combines several dimensions.<\/p>\n\n\n\n<h3 id=\"aioseo-token-usage\" class=\"wp-block-heading\">Token Usage<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/wrangleai.com\/blog\/how-ai-cost-optimisation-software-stops-token-waste\/\">Tokens remain one of the most important measurements<\/a> because model pricing commonly depends on the amount of information processed and generated.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Teams should understand input and output tokens, while also accounting for provider-specific categories such as cached tokens or reasoning-related usage where relevant.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI, for example, exposes token usage information through API responses and its Usage Dashboard. Depending on the endpoint and model, additional details can include cached-input and reasoning-token counts.<\/p>\n\n\n\n<h3 id=\"aioseo-request-volume\" class=\"wp-block-heading\">Request Volume<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Teams should monitor how many requests each workload generates.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A sharp increase may indicate legitimate growth, but it could also indicate an application bug, automated loop, unusual user activity or compromised credential.<\/p>\n\n\n\n<h3 id=\"aioseo-cost\" class=\"wp-block-heading\">Cost<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Token counts are useful, but engineering leaders also need to understand financial impact.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Different models have different prices, and the same workload can have very different costs depending on which model processes it.<\/p>\n\n\n\n<h3 id=\"aioseo-model-usage\" class=\"wp-block-heading\">Model Usage<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Teams should know which models are being used and how that changes over time.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This becomes particularly important when several providers are available to developers.<\/p>\n\n\n\n<h3 id=\"aioseo-application-and-agent-usage\" class=\"wp-block-heading\">Application and Agent Usage<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Provider-level totals do not provide enough accountability.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Organisations should try to understand which applications, projects and agents are responsible for consumption.<\/p>\n\n\n\n<h2 id=\"aioseo-how-to-track-openai-api-usage\" class=\"wp-block-heading\">How to Track OpenAI API Usage<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/openai.com\/\">OpenAI<\/a> provides several ways to review API usage.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Its Usage Dashboard allows authorised users to review activity across billing periods and filter usage by projects. Usage information can also be obtained directly from API responses for individual requests.<\/p>\n\n\n\n<h3 id=\"aioseo-use-the-openai-usage-dashboard\" class=\"wp-block-heading\">Use the OpenAI Usage Dashboard<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The Usage Dashboard provides a useful starting point for organisations using OpenAI APIs.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Teams can select reporting periods and filter activity by project. OpenAI also allows users with the appropriate permissions to review specific API capability usage and inspect more detailed activity.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This is useful when teams have structured their OpenAI environment around clear projects.<\/p>\n\n\n\n<h3 id=\"aioseo-track-tokens-from-api-responses\" class=\"wp-block-heading\">Track Tokens from API Responses<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI API responses can include usage information.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For example, depending on the endpoint, teams can obtain input, output and total token counts. Some endpoints and models provide further token categories that can help explain the true amount of processing behind a request.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Capturing this information inside application telemetry can provide more detailed workload-level monitoring.<\/p>\n\n\n\n<h3 id=\"aioseo-export-openai-usage-data\" class=\"wp-block-heading\">Export OpenAI Usage Data<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI also supports exporting detailed usage and cost information.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Activity data can be grouped by dimensions including project, user, API key, model, batch or service tier, while cost exports can support financial reporting and invoice reconciliation.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This can be useful for internal reporting, but it still represents the OpenAI part of the wider AI environment.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><a href=\"https:\/\/wrangleai.com\/demo\/\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1024\" height=\"171\" src=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-1024x171.png\" alt=\"CTA\" class=\"wp-image-272\" srcset=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-1024x171.png 1024w, https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-300x50.png 300w, https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-768x128.png 768w, https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2.png 1200w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/a><\/figure>\n\n\n\n<h2 id=\"aioseo-how-to-track-anthropic-claude-usage\" class=\"wp-block-heading\">How to Track Anthropic Claude Usage<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.anthropic.com\/\">Anthropic<\/a> provides usage and cost reporting through the Claude Console.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The Console can show API usage across workspaces, models and API keys, including input and output token information. It also provides cost reporting for supported users.<\/p>\n\n\n\n<h3 id=\"aioseo-monitor-usage-by-model\" class=\"wp-block-heading\">Monitor Usage by Model<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Engineering teams can filter Anthropic usage by model.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This helps organisations understand which Claude models are responsible for consumption and how model usage changes over time.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Model-level visibility is important because choosing a different model can significantly change both performance and cost.<\/p>\n\n\n\n<h3 id=\"aioseo-monitor-usage-by-api-key\" class=\"wp-block-heading\">Monitor Usage by API Key<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Anthropic&#8217;s reporting can also break down usage by API key.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This is useful when organisations issue separate keys to different applications, teams or workloads.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A good key structure improves attribution. If every application shares the same key, the organisation loses much of the operational value that provider reporting can offer.<\/p>\n\n\n\n<h3 id=\"aioseo-track-input-and-output-tokens\" class=\"wp-block-heading\">Track Input and Output Tokens<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The Claude Console reports input and output token usage and provides more detailed views at smaller time intervals. Teams can also review rate-limited requests and compare token activity with applicable rate limits.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This can help engineering teams understand both consumption and capacity pressure.<\/p>\n\n\n\n<h3 id=\"aioseo-export-anthropic-usage\" class=\"wp-block-heading\">Export Anthropic Usage<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Anthropic supports CSV exports from its usage and cost reports. This allows teams to analyse the information outside the provider console or combine it with other internal reporting.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For organisations that need programmatic reporting, Anthropic also provides administrative usage reporting capabilities that can group usage by dimensions such as API key, workspace and model.<\/p>\n\n\n\n<h2 id=\"aioseo-how-to-track-google-gemini-usage\" class=\"wp-block-heading\">How to Track Google Gemini Usage<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/ai.google.dev\/gemini-api\/\">Google Gemini API<\/a> provides usage monitoring through Google AI Studio and Google&#8217;s billing infrastructure.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Google states that Gemini API users can review usage through AI Studio under Dashboard &gt; Usage. Paid usage is connected with Cloud Billing, while billing status and spending information can also be managed through the relevant Google environment.<\/p>\n\n\n\n<h3 id=\"aioseo-track-gemini-usage-through-ai-studio\" class=\"wp-block-heading\">Track Gemini Usage Through AI Studio<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Google AI Studio provides the most direct view for many Gemini API developers.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Teams can use its usage dashboard to understand API consumption associated with their projects. Every Gemini API key is associated with a Google Cloud project, which makes project structure important for monitoring and governance.<\/p>\n\n\n\n<h3 id=\"aioseo-monitor-gemini-costs\" class=\"wp-block-heading\">Monitor Gemini Costs<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Gemini API billing can depend on input tokens, output tokens, cached tokens, cache storage and other model or tool-specific usage. This makes simple request counting insufficient for understanding total cost.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Teams should therefore monitor both consumption and financial data.<\/p>\n\n\n\n<h3 id=\"aioseo-use-gemini-api-logs-where-appropriate\" class=\"wp-block-heading\">Use Gemini API Logs Where Appropriate<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Google has also introduced Gemini API logging capabilities for supported paid-tier projects. These logs can help developers understand model behaviour and application interactions, although organisations should carefully consider what data they store and their own privacy and governance requirements.<\/p>\n\n\n\n<h2 id=\"aioseo-the-problem-with-tracking-each-provider-separately\" class=\"wp-block-heading\">The Problem with Tracking Each Provider Separately<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI, Anthropic and Google each provide useful monitoring capabilities.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The problem is not that provider dashboards are poor. The problem is that each dashboard sees only its own environment.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Imagine an organisation with 30 AI-powered applications.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Ten use OpenAI. Eight use Claude. Seven use Gemini. The remaining applications dynamically switch between providers.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Engineering leadership asks a simple question:<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Which product generated the most AI cost this month?<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The answer may require exporting data from several systems, normalising different formats, mapping API keys back to applications, comparing different pricing structures and manually combining the results.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That process becomes increasingly difficult as AI adoption grows.<\/p>\n\n\n\n<h2 id=\"aioseo-create-a-common-ai-usage-data-model\" class=\"wp-block-heading\">Create a Common AI Usage Data Model<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Multi-provider organisations need a common way to describe AI consumption.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Instead of organising everything only around provider terminology, teams should create business-level dimensions that work across all platforms.<\/p>\n\n\n\n<h3 id=\"aioseo-provider\" class=\"wp-block-heading\">Provider<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Identify whether the request went to OpenAI, Anthropic, Gemini or another approved provider.<\/p>\n\n\n\n<h3 id=\"aioseo-model\" class=\"wp-block-heading\">Model<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Record the actual model that processed the request.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This is particularly important when applications can switch models.<\/p>\n\n\n\n<h3 id=\"aioseo-application\" class=\"wp-block-heading\">Application<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Connect usage with the product or service responsible for generating it.<\/p>\n\n\n\n<h3 id=\"aioseo-team\" class=\"wp-block-heading\">Team<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Identify the team responsible for the workload.<\/p>\n\n\n\n<h3 id=\"aioseo-api-key-or-identity\" class=\"wp-block-heading\">API Key or Identity<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Track the technical identity responsible for the request where appropriate.<\/p>\n\n\n\n<h3 id=\"aioseo-agent\" class=\"wp-block-heading\">Agent<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">For agentic systems, identify which agent generated the activity.<\/p>\n\n\n\n<h3 id=\"aioseo-cost-2\" class=\"wp-block-heading\">Cost<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Normalise costs so teams can compare workloads across providers.<\/p>\n\n\n\n<h3 id=\"aioseo-time\" class=\"wp-block-heading\">Time<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Record when usage occurred so teams can understand trends and identify unusual activity.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This common structure makes cross-provider reporting far more useful.<\/p>\n\n\n\n<h2 id=\"aioseo-use-separate-keys-and-identities-for-better-attribution\" class=\"wp-block-heading\">Use Separate Keys and Identities for Better Attribution<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">One of the simplest ways to improve AI usage monitoring is to stop sharing credentials across unrelated workloads.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Separate keys or identities can be assigned to individual applications, teams, projects or agents.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This provides much clearer attribution.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If an application suddenly generates unusual activity, engineering teams can identify the responsible workload rather than seeing only an organisation-wide increase.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Key-level separation also creates opportunities for more targeted governance controls.<\/p>\n\n\n\n<h2 id=\"aioseo-monitor-ai-usage-by-application-not-just-provider\" class=\"wp-block-heading\">Monitor AI Usage by Application, Not Just Provider<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Provider-level reporting answers the question, &#8220;How much are we using OpenAI?&#8221;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Business leaders usually need a different answer:<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">&#8220;Which products are driving our AI spending?&#8221;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This distinction matters.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">An organisation may spend 50 per cent of its AI budget with one provider, but that information does not explain which application creates the cost or whether the spending delivers enough business value.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Application-level monitoring allows teams to compare usage with product activity.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This creates a stronger foundation for AI FinOps and governance.<\/p>\n\n\n\n<h2 id=\"aioseo-track-ai-usage-by-agent\" class=\"wp-block-heading\">Track AI Usage by Agent<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Agentic AI requires more detailed monitoring.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">An agent can make many model calls during a single task, and those calls may even move across providers.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For example, an agent could use one model for planning, another for analysing a document and a third as a fallback.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">From the user&#8217;s perspective, this may appear to be one action.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">From the infrastructure perspective, it could involve many requests and a much larger token footprint.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Teams should therefore attribute <strong>AI usage<\/strong> to individual agents where possible.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This makes it easier to identify inefficient agents, unexpected loops and unusual spending.<\/p>\n\n\n\n<h2 id=\"aioseo-monitor-ai-costs-across-providers\" class=\"wp-block-heading\">Monitor AI Costs Across Providers<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Cost tracking becomes more difficult when models use different pricing structures.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The goal should be to convert provider-specific billing into a common operational view.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Engineering teams should be able to compare:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Cost by provider<\/li>\n\n\n\n<li>Cost by model<\/li>\n\n\n\n<li>Cost by application<\/li>\n\n\n\n<li>Cost by project<\/li>\n\n\n\n<li>Cost by API key<\/li>\n\n\n\n<li>Cost by agent<\/li>\n\n\n\n<li>Cost over time<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">This allows organisations to move from invoice tracking towards active AI cost management.<\/p>\n\n\n\n<h2 id=\"aioseo-build-alerts-for-unexpected-ai-usage\" class=\"wp-block-heading\">Build Alerts for Unexpected AI Usage<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Dashboards explain what has happened. Alerts help teams react while it is happening.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Organisations should establish normal usage patterns and create alerts for meaningful changes.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Examples can include unexpected token growth, sudden increases in request volume, unusual activity from a specific key, rapid agent consumption, abnormal spending or unexpected model changes.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Thresholds should reflect each workload.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A production customer assistant will naturally have different usage patterns from an internal development tool.<\/p>\n\n\n\n<h2 id=\"aioseo-track-which-model-actually-processes-each-request\" class=\"wp-block-heading\">Track Which Model Actually Processes Each Request<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Multi-model architectures introduce another important monitoring requirement.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Applications may route requests dynamically based on cost, availability, latency or task complexity.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This means the model requested by an application may not always be the model that ultimately handles the workload.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Teams should therefore monitor routing outcomes.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This helps answer questions such as whether expensive models are being overused, whether fallback models are activated frequently, and whether workloads remain within approved provider policies.<\/p>\n\n\n\n<h2 id=\"aioseo-connect-ai-usage-monitoring-with-governance\" class=\"wp-block-heading\">Connect AI Usage Monitoring with Governance<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">AI monitoring should support more than cost management.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">It can also strengthen governance.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Central visibility helps organisations understand which models are active, whether approved providers are being used, who owns each workload, and whether unusual activity requires investigation.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This becomes especially important for organisations working towards frameworks such as ISO\/IEC 42001 or NIST AI RMF, or managing obligations under regulations such as the EU AI Act.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Monitoring does not automatically create compliance, but strong operational visibility can provide useful evidence and accountability within a wider governance programme.<\/p>\n\n\n\n<h2 id=\"aioseo-create-a-central-ai-usage-dashboard\" class=\"wp-block-heading\">Create a Central AI Usage Dashboard<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">For organisations operating across OpenAI, Anthropic and Gemini, the long-term goal should be a central operational view.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The dashboard should allow engineering, FinOps, governance and leadership teams to understand AI activity without manually switching between provider consoles.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Useful views can include total AI spending, provider distribution, model usage, application costs, key-level activity, agent consumption, token trends and unusual usage events.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The purpose is not to replace every provider dashboard.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Provider tools remain useful for detailed platform-specific investigation.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The central layer should provide the organisation-wide view that individual providers cannot.<\/p>\n\n\n\n<h2 id=\"aioseo-how-wrangleai-helps-track-ai-usage-across-providers\" class=\"wp-block-heading\">How WrangleAI Helps Track AI Usage Across Providers<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/wrangleai.com\/\">WrangleAI<\/a> helps organisations solve the visibility problem created by multi-provider AI infrastructure.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Instead of treating OpenAI, Anthropic and Gemini as separate reporting environments, WrangleAI gives teams a more central operational layer for understanding <strong>AI usage<\/strong> across providers, models, keys, projects, applications and agents.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This helps engineering teams see where AI resources are being consumed, understand costs at a more useful level and identify unexpected changes before they become larger problems.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">WrangleAI also supports the wider controls organisations need as AI infrastructure matures, including budget management, <a href=\"https:\/\/wrangleai.com\/blog\/checklist-for-ai-cost-visibility\/\">AI key visibility<\/a>, auditability and intelligent model routing.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">SmartRouter can help teams move workloads between suitable models based on factors such as cost, performance and availability. When routing and usage visibility work together, organisations can understand not only how much AI they are consuming, but also where requests are going and whether model selection remains efficient.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This creates a stronger foundation for AI FinOps, governance and multi-provider operations.<\/p>\n\n\n\n<h2 id=\"aioseo-final-thoughts\" class=\"wp-block-heading\">Final Thoughts<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Tracking AI usage is simple when an organisation has one model, one API key and one application.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That simplicity disappears when AI reaches production at scale.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI, Anthropic and Gemini each provide useful tools for monitoring their own platforms. OpenAI offers organisation and project-level usage reporting and detailed exports. Anthropic provides reporting across models, workspaces and API keys. Google provides Gemini API usage and billing visibility through AI Studio and its wider cloud environment.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The challenge is connecting those separate views.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Organisations need to understand AI consumption by provider, model, application, team, key and agent. They need to connect token activity with costs, monitor unusual behaviour and understand which models actually process workloads.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">WrangleAI helps create this central operational layer.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">By bringing AI usage visibility, cost monitoring, key-level attribution, model routing, budget controls and governance together, WrangleAI helps engineering teams manage multi-provider AI without losing control as their infrastructure grows.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The goal is not simply to know how many tokens were consumed.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The goal is to understand where AI is being used, what it costs, who is responsible for it and whether that usage supports the organisation&#8217;s wider business and governance goals.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><a href=\"https:\/\/wrangleai.com\/demo\/\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1024\" height=\"171\" src=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-1024x171.png\" alt=\"CTA\" class=\"wp-image-272\" srcset=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-1024x171.png 1024w, https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-300x50.png 300w, https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2-768x128.png 768w, https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/09\/WrangleAI-CTA-2.png 1200w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/a><\/figure>\n\n\n\n<h2 id=\"aioseo-faqs\" class=\"wp-block-heading\">FAQs<\/h2>\n\n\n\n<div id=\"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\" class=\"wp-block-aioseo-faq\" data-schema-only=\"false\"><h3 class=\"aioseo-faq-block-question\">How can I track AI usage across OpenAI, Anthropic and Gemini?<\/h3><div class=\"aioseo-faq-block-answer\">\n<p class=\"wp-block-paragraph\">Each provider offers its own usage and billing tools. OpenAI provides an API Usage Dashboard, Anthropic provides usage and cost reporting through the Claude Console, and Gemini API usage can be monitored through Google AI Studio. Organisations using all three should also consider a central monitoring layer to combine provider-level data into a consistent view.<\/p>\n<\/div><\/div>\n\n\n\n<div id=\"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\" class=\"wp-block-aioseo-faq\" data-schema-only=\"false\"><h3 class=\"aioseo-faq-block-question\">What AI usage metrics should engineering teams track?<\/h3><div class=\"aioseo-faq-block-answer\">\n<p class=\"wp-block-paragraph\">Engineering teams should track input and output tokens, request volume, costs, provider, model, application, API key or identity, agent activity and usage trends. Multi-model environments should also track routing decisions so teams know which model actually processes each workload.<\/p>\n<\/div><\/div>\n\n\n\n<div id=\"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\" class=\"wp-block-aioseo-faq\" data-schema-only=\"false\"><h3 class=\"aioseo-faq-block-question\">Why are provider dashboards not enough for multi-provider AI?<\/h3><div class=\"aioseo-faq-block-answer\">\n<p class=\"wp-block-paragraph\">Provider dashboards are designed to monitor their own ecosystems. If an organisation uses several providers, the data remains separated. A central monitoring layer makes it easier to compare usage, costs, models, applications and agents across the complete AI environment.<\/p>\n<\/div><\/div>\n\n\n\n<div id=\"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\" class=\"wp-block-aioseo-faq\" data-schema-only=\"false\"><h3 class=\"aioseo-faq-block-question\">How does WrangleAI help track AI usage?<\/h3><div class=\"aioseo-faq-block-answer\">\n<p class=\"wp-block-paragraph\">WrangleAI provides a central operational layer for monitoring AI usage across providers, models, keys, applications, projects and agents. It can help teams improve cost visibility, workload attribution, budget control, auditability and model routing across multi-provider AI environments.<\/p>\n<\/div><\/div>\n\n\n\n<div id=\"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini\" class=\"wp-block-aioseo-faq\" data-schema-only=\"false\"><h3 class=\"aioseo-faq-block-question\">Can AI usage monitoring help reduce AI costs?<\/h3><div class=\"aioseo-faq-block-answer\">\n<p class=\"wp-block-paragraph\">Yes. Detailed usage monitoring helps teams identify expensive models, inefficient applications, unusually large prompts, unexpected request growth and agent workflows that generate excessive calls. Teams can then optimise workloads or route suitable tasks to more efficient models.<\/p>\n<\/div><\/div>\n","protected":false},"excerpt":{"rendered":"<p>Enterprise AI infrastructure is changing quickly. Many organisations no longer rely on a single model provider. One engineering team may [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":205,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[5,6],"tags":[],"class_list":["post-427","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-cost-allocation","category-ai-performance-optimisation"],"aioseo_notices":[],"aioseo_head":"\n\t\t<!-- All in One SEO 5.0.1.1 - aioseo.com -->\n\t<meta name=\"description\" content=\"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.\" \/>\n\t<meta name=\"robots\" content=\"max-image-preview:large\" \/>\n\t<meta name=\"author\" content=\"admin\"\/>\n\t<link rel=\"canonical\" href=\"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/\" \/>\n\t<meta name=\"generator\" content=\"All in One SEO (AIOSEO) 5.0.1.1\" \/>\n\t\t<meta property=\"og:locale\" content=\"en_US\" \/>\n\t\t<meta property=\"og:site_name\" content=\"|\" \/>\n\t\t<meta property=\"og:type\" content=\"article\" \/>\n\t\t<meta property=\"og:title\" content=\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\" \/>\n\t\t<meta property=\"og:description\" content=\"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.\" \/>\n\t\t<meta property=\"og:url\" content=\"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/\" \/>\n\t\t<meta property=\"og:image\" content=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png\" \/>\n\t\t<meta property=\"og:image:secure_url\" content=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png\" \/>\n\t\t<meta property=\"og:image:width\" content=\"820\" \/>\n\t\t<meta property=\"og:image:height\" content=\"447\" \/>\n\t\t<meta property=\"article:published_time\" content=\"2026-09-01T10:49:40+00:00\" \/>\n\t\t<meta property=\"article:modified_time\" content=\"2026-09-01T10:49:42+00:00\" \/>\n\t\t<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n\t\t<meta name=\"twitter:title\" content=\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\" \/>\n\t\t<meta name=\"twitter:description\" content=\"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.\" \/>\n\t\t<meta name=\"twitter:image\" content=\"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png\" \/>\n\t\t<script type=\"application\/ld+json\" class=\"aioseo-schema\">\n\t\t\t{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"BlogPosting\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#blogposting\",\"name\":\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\",\"headline\":\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\",\"author\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/author\\\/admin_wrangle\\\/#author\"},\"publisher\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/#organization\"},\"image\":{\"@type\":\"ImageObject\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/wp-content\\\/uploads\\\/2025\\\/07\\\/AI-API-usage.png\",\"width\":820,\"height\":447,\"caption\":\"AI API usage\"},\"datePublished\":\"2026-09-01T10:49:40+00:00\",\"dateModified\":\"2026-09-01T10:49:42+00:00\",\"inLanguage\":\"en-US\",\"commentCount\":1,\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#webpage\"},\"isPartOf\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#webpage\"},\"articleSection\":\"AI Cost Allocation, AI Performance Optimisation\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#breadcrumblist\",\"itemListElement\":[{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog#listItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/wrangleai.com\\\/blog\",\"nextItem\":{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/category\\\/ai-cost-allocation\\\/#listItem\",\"name\":\"AI Cost Allocation\"}},{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/category\\\/ai-cost-allocation\\\/#listItem\",\"position\":2,\"name\":\"AI Cost Allocation\",\"item\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/category\\\/ai-cost-allocation\\\/\",\"nextItem\":{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#listItem\",\"name\":\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\"},\"previousItem\":{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog#listItem\",\"name\":\"Home\"}},{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#listItem\",\"position\":3,\"name\":\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\",\"previousItem\":{\"@type\":\"ListItem\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/category\\\/ai-cost-allocation\\\/#listItem\",\"name\":\"AI Cost Allocation\"}}]},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/#organization\",\"name\":\"WrangleAI\",\"description\":\"Monitor, optimize, and control token usage across every model, provider, and team.\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/wp-content\\\/uploads\\\/2025\\\/06\\\/main_1.png\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#organizationLogo\",\"width\":1024,\"height\":1024,\"caption\":\"Wrangle AI logo\"},\"image\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#organizationLogo\"}},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/author\\\/admin_wrangle\\\/#author\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/author\\\/admin_wrangle\\\/\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#authorImage\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/97d5421b31b2eae711ac75cf15198600b6f89d890abf3990c0696d3b3d86a1de?s=96&d=mm&r=g\",\"width\":96,\"height\":96,\"caption\":\"admin\"}},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#webpage\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/\",\"name\":\"How to Track AI Usage Across OpenAI, Anthropic and Gemini\",\"description\":\"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.\",\"inLanguage\":\"en-US\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/#website\"},\"breadcrumb\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#breadcrumblist\"},\"author\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/author\\\/admin_wrangle\\\/#author\"},\"creator\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/author\\\/admin_wrangle\\\/#author\"},\"image\":{\"@type\":\"ImageObject\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/wp-content\\\/uploads\\\/2025\\\/07\\\/AI-API-usage.png\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#mainImage\",\"width\":820,\"height\":447,\"caption\":\"AI API usage\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\\\/#mainImage\"},\"datePublished\":\"2026-09-01T10:49:40+00:00\",\"dateModified\":\"2026-09-01T10:49:42+00:00\"},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/#website\",\"url\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/\",\"name\":\"Get a Grip on your AI Stack with WrangleAI\",\"alternateName\":\"WrangleAI\",\"inLanguage\":\"en-US\",\"publisher\":{\"@id\":\"https:\\\/\\\/wrangleai.com\\\/blog\\\/#organization\"}}]}\n\t\t<\/script>\n\t\t<!-- All in One SEO -->\n\n","aioseo_head_json":{"title":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","description":"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.","canonical_url":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"BlogPosting","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#blogposting","name":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","headline":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","author":{"@id":"https:\/\/wrangleai.com\/blog\/author\/admin_wrangle\/#author"},"publisher":{"@id":"https:\/\/wrangleai.com\/blog\/#organization"},"image":{"@type":"ImageObject","url":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png","width":820,"height":447,"caption":"AI API usage"},"datePublished":"2026-09-01T10:49:40+00:00","dateModified":"2026-09-01T10:49:42+00:00","inLanguage":"en-US","commentCount":1,"mainEntityOfPage":{"@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#webpage"},"isPartOf":{"@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#webpage"},"articleSection":"AI Cost Allocation, AI Performance Optimisation"},{"@type":"BreadcrumbList","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#breadcrumblist","itemListElement":[{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog#listItem","position":1,"name":"Home","item":"https:\/\/wrangleai.com\/blog","nextItem":{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog\/category\/ai-cost-allocation\/#listItem","name":"AI Cost Allocation"}},{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog\/category\/ai-cost-allocation\/#listItem","position":2,"name":"AI Cost Allocation","item":"https:\/\/wrangleai.com\/blog\/category\/ai-cost-allocation\/","nextItem":{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#listItem","name":"How to Track AI Usage Across OpenAI, Anthropic and Gemini"},"previousItem":{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog#listItem","name":"Home"}},{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#listItem","position":3,"name":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","previousItem":{"@type":"ListItem","@id":"https:\/\/wrangleai.com\/blog\/category\/ai-cost-allocation\/#listItem","name":"AI Cost Allocation"}}]},{"@type":"Organization","@id":"https:\/\/wrangleai.com\/blog\/#organization","name":"WrangleAI","description":"Monitor, optimize, and control token usage across every model, provider, and team.","url":"https:\/\/wrangleai.com\/blog\/","logo":{"@type":"ImageObject","url":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/06\/main_1.png","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#organizationLogo","width":1024,"height":1024,"caption":"Wrangle AI logo"},"image":{"@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#organizationLogo"}},{"@type":"Person","@id":"https:\/\/wrangleai.com\/blog\/author\/admin_wrangle\/#author","url":"https:\/\/wrangleai.com\/blog\/author\/admin_wrangle\/","name":"admin","image":{"@type":"ImageObject","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#authorImage","url":"https:\/\/secure.gravatar.com\/avatar\/97d5421b31b2eae711ac75cf15198600b6f89d890abf3990c0696d3b3d86a1de?s=96&d=mm&r=g","width":96,"height":96,"caption":"admin"}},{"@type":"WebPage","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#webpage","url":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/","name":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","description":"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.","inLanguage":"en-US","isPartOf":{"@id":"https:\/\/wrangleai.com\/blog\/#website"},"breadcrumb":{"@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#breadcrumblist"},"author":{"@id":"https:\/\/wrangleai.com\/blog\/author\/admin_wrangle\/#author"},"creator":{"@id":"https:\/\/wrangleai.com\/blog\/author\/admin_wrangle\/#author"},"image":{"@type":"ImageObject","url":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png","@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#mainImage","width":820,"height":447,"caption":"AI API usage"},"primaryImageOfPage":{"@id":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/#mainImage"},"datePublished":"2026-09-01T10:49:40+00:00","dateModified":"2026-09-01T10:49:42+00:00"},{"@type":"WebSite","@id":"https:\/\/wrangleai.com\/blog\/#website","url":"https:\/\/wrangleai.com\/blog\/","name":"Get a Grip on your AI Stack with WrangleAI","alternateName":"WrangleAI","inLanguage":"en-US","publisher":{"@id":"https:\/\/wrangleai.com\/blog\/#organization"}}]},"og:locale":"en_US","og:site_name":"|","og:type":"article","og:title":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","og:description":"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.","og:url":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/","og:image":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png","og:image:secure_url":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png","og:image:width":"820","og:image:height":"447","article:published_time":"2026-09-01T10:49:40+00:00","article:modified_time":"2026-09-01T10:49:42+00:00","twitter:card":"summary_large_image","twitter:title":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","twitter:description":"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.","twitter:image":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png"},"aioseo_meta_data":{"post_id":"427","title":"#post_title","description":"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.","keywords":null,"keyphrases":{"focus":{"keyphrase":"AI Usage","score":0,"analysis":[]},"additional":[]},"primary_term":null,"canonical_url":null,"og_title":"#post_title","og_description":"Learn how to track AI usage across OpenAI, Anthropic and Gemini, including tokens, costs, models, API keys, applications, agents and central monitoring.","og_object_type":"default","og_image_type":"featured","og_image_url":"https:\/\/wrangleai.com\/blog\/wp-content\/uploads\/2025\/07\/AI-API-usage.png","og_image_width":"820","og_image_height":"447","og_image_custom_url":null,"og_image_custom_fields":null,"og_video":"","og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":true,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[{"hidden":false,"schemaBlockId":"aioseo-mtijajyphf2rw","tagName":"h3","answer":"Each provider offers its own usage and billing tools. OpenAI provides an API Usage Dashboard, Anthropic provides usage and cost reporting through the Claude Console, and Gemini API usage can be monitored through Google AI Studio. Organisations using all three should also consider a central monitoring layer to combine provider-level data into a consistent view.","question":"How can I track AI usage across OpenAI, Anthropic and Gemini?","anchor":"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini","type":"aioseo\/faq"},{"hidden":false,"question":"What AI usage metrics should engineering teams track?","schemaBlockId":"aioseo-mtijawkxx588e","tagName":"h3","anchor":"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini","answer":"Engineering teams should track input and output tokens, request volume, costs, provider, model, application, API key or identity, agent activity and usage trends. Multi-model environments should also track routing decisions so teams know which model actually processes each workload.","type":"aioseo\/faq"},{"hidden":false,"question":"Why are provider dashboards not enough for multi-provider AI?","schemaBlockId":"aioseo-mtijba52pta0j","tagName":"h3","anchor":"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini","answer":"Provider dashboards are designed to monitor their own ecosystems. If an organisation uses several providers, the data remains separated. A central monitoring layer makes it easier to compare usage, costs, models, applications and agents across the complete AI environment.","type":"aioseo\/faq"},{"hidden":false,"question":"How does WrangleAI help track AI usage?","schemaBlockId":"aioseo-mtijbmvj1up5h","tagName":"h3","anchor":"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini","answer":"WrangleAI provides a central operational layer for monitoring AI usage across providers, models, keys, applications, projects and agents. It can help teams improve cost visibility, workload attribution, budget control, auditability and model routing across multi-provider AI environments.","type":"aioseo\/faq"},{"hidden":false,"question":"Can AI usage monitoring help reduce AI costs?","schemaBlockId":"aioseo-mtijbzf9cuci3","tagName":"h3","anchor":"aioseo-how-can-i-track-ai-usage-across-openai-anthropic-and-gemini","answer":"Yes. Detailed usage monitoring helps teams identify expensive models, inefficient applications, unusually large prompts, unexpected request growth and agent workflows that generate excessive calls. Teams can then optimise workloads or route suitable tasks to more efficient models.","type":"aioseo\/faq"}],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"BlogPosting","isEnabled":true},"graphs":[]},"schema_type":"default","schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":"-1","robots_max_videopreview":"-1","robots_max_imagepreview":"large","priority":null,"frequency":"default","local_seo":null,"breadcrumb_settings":null,"limit_modified_date":false,"ai":{"faqs":[],"keyPoints":[],"schemas":[],"titles":[],"descriptions":[],"socialPosts":{"email":{"subject":"","preview":"","content":""},"linkedin":[],"twitter":[],"facebook":[],"instagram":[]}},"created":"2026-09-01 10:49:42","updated":"2026-09-01 10:52:31","seo_analyzer_scan_date":null,"focus_keyword":"AI Usage","additional_keywords":null,"truseo_locale":null},"aioseo_breadcrumb":"<div class=\"aioseo-breadcrumbs\"><span class=\"aioseo-breadcrumb\">\n\t\t\t<a href=\"https:\/\/wrangleai.com\/blog\" title=\"Home\">Home<\/a>\n\t\t<\/span><span class=\"aioseo-breadcrumb-separator\">&raquo;<\/span><span class=\"aioseo-breadcrumb\">\n\t\t\t<a href=\"https:\/\/wrangleai.com\/blog\/category\/ai-cost-allocation\/\" title=\"AI Cost Allocation\">AI Cost Allocation<\/a>\n\t\t<\/span><span class=\"aioseo-breadcrumb-separator\">&raquo;<\/span><span class=\"aioseo-breadcrumb\">\n\t\t\tHow to Track AI Usage Across OpenAI, Anthropic and Gemini\n\t\t<\/span><\/div>","aioseo_breadcrumb_json":[{"label":"Home","link":"https:\/\/wrangleai.com\/blog"},{"label":"AI Cost Allocation","link":"https:\/\/wrangleai.com\/blog\/category\/ai-cost-allocation\/"},{"label":"How to Track AI Usage Across OpenAI, Anthropic and Gemini","link":"https:\/\/wrangleai.com\/blog\/how-to-track-ai-usage-across-openai-anthropic-and-gemini\/"}],"_links":{"self":[{"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/posts\/427","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/comments?post=427"}],"version-history":[{"count":1,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/posts\/427\/revisions"}],"predecessor-version":[{"id":428,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/posts\/427\/revisions\/428"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/media\/205"}],"wp:attachment":[{"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/media?parent=427"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/categories?post=427"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/wrangleai.com\/blog\/wp-json\/wp\/v2\/tags?post=427"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}