How to Use a Skill in Claude: App, Claude Code, and API
Learn how to select a Skill in Claude, add and test a custom Skill in Claude’s app or Claude Code, and understand what the Claude API supports.
AI model and infrastructure insights
Learn how to select a Skill in Claude, add and test a custom Skill in Claude’s app or Claude Code, and understand what the Claude API supports.
Troubleshoot Gemini API 429 errors by matching the deployed key to its project, checking active model limits and tier in AI Studio, and choosing a product-specific recovery step.
Connect to GPT-5.5 through OpenAI’s API, make a first Responses API request, check for completed output, and add schema-constrained data to an application.
Choose Gemini for personal use or software integration. Follow Android setup steps or prepare a Python environment, create an API key, and send a documented test request.
Open a blank Canvas in ChatGPT from a desktop browser, see what to try if it is missing, and learn how Canvas differs from the production API.
See reported Gemini 2.5 Pro API token rates, sample cost calculations, what to include for thinking tokens, and the current access and routing qualifications.
Claude Code’s documented gateways do not support routing to Gemini models. Learn how to use Gemini CLI separately or call the Gemini API from your application.
Is a Grok 4 free trial available? See what xAI lists for free and paid access, how Grok 4.7 API pricing works, and which trial and billing terms remain unconfirmed.
The Claude API isn’t generally free for ordinary use. See current USD token prices, who may qualify for API credits, and how to estimate a production budget.
We could not verify a fixed image count in the official guidance reviewed. Learn what to do at the limit and how separate API image costs work for production.
Add an MCP server to Claude Code with a documented command, choose local, project, or user scope, and check its connection. Learn where project configuration belongs and what to check if a server is missing.
Yes, Opus 4.6 is selectable in Claude Code if it’s available to your account. See the selection commands, Anthropic API setup and pricing, and how to evaluate it against the current Opus model.
Gemini app limits depend on usage, while API free access has separate model- and project-specific limits. See what to check before budgeting API use.
Update Gemini CLI using the method that matches your installation, verify the version, and troubleshoot failures without starting with destructive cleanup.
Claude’s Free plan doesn’t include Opus. See Opus 4.5 API pricing, how paid-plan credits differ from free access, and what to check before a production deployment.
Get started with OpenAI Codex: install the desktop app, open a project, ask for a focused code change, review and test the result, and see how Codex differs from the OpenAI API.
Suno’s free plan provides 50 credits daily, with UTC replenishment and an official estimate of up to 10 songs. Compare annual plans, download limits, and API options.
Learn when to use a Claude Skill, create and test a minimal SKILL.md in Claude, and use a built-in Excel Skill through the Messages API—including how to download and check its generated workbook.
Compare Ultracode workflows with max effort, distinguish both from the Claude Max subscription, and see what the documented API settings do—and don’t—confirm.
OpenAI discontinued the Sora API on September 24, 2026. Here’s what that means for current pricing and production budgets.
Learn how to make a minimal Python request to GPT-5.2 through OpenAI’s API, and review its documented model IDs, endpoints, limits, pricing, and lifecycle notes.
See the GPT-5.5 rates reported by an Inworld model listing, calculate an example API budget, and learn why those figures should not be assumed to be direct OpenAI API prices.
Find out what free Claude Opus 4.7 access is confirmed, what API use costs, and how to estimate a bill before integrating it in production.
Compare OpenAI embedding prices for standard and Batch processing, calculate a token-based budget, and see when Batch may fit your workload.