API Guides

Step-by-step AI API integration: getting and securing API keys, authentication, endpoints and SDKs, OpenAI-compatible setups, and patterns for production use.

62 articles
Updated Oct 10, 2026

Conceptual Claude API key ownership guide: create and revoke your own key, track billing, and match the endpoint; Max and Team offer claimed monthly API credits, while Pro includes none.

Latest

Claude API Key: How to Get One, Is It Free, and Safe Buying

Creating a Claude API key is free; calls spend prepaid credits. New users get a small test credit, Max and Team add monthly API credits, and Pro adds none.

Oct 10, 202616 min

Free image API allowances as of October 10, 2026: about 173 daily FLUX images through Cloudflare Workers AI, about 250 ModelScope Qwen Image 2.1 calls for verified non-commercial use, and no Hugging Face monthly credit on free accounts.

Free AI Image Generation API in 2026: Daily Image Limits Compared

As of October 10, 2026, Cloudflare Workers AI is the main recurring free image API: about 173 FLUX images a day. Hugging Face, Gemini and OpenAI give none.

Oct 10, 202614 min

Free image API guide comparing Unsplash, Pexels, Pixabay, and Wikimedia Commons

Free Image API in 2026: Unsplash 1,000/hr, Pexels New Keys Paused

As of October 10, 2026, Unsplash allows 1,000 requests an hour after approval, Pexels has paused new API keys, and Pixabay or keyless Openverse work today.

Oct 10, 202618 min

Gemini API access map connecting an official project key to billing, usage, and a separate provider contract

Gemini API Pricing 2026: Token Rates, Free Tier, and Key Cost

A Gemini API key is free, but usage is free only within per-project limits. Gemini 3.8 Flash costs $0.75/$3.75 per 1M tokens, doubling on January 1, 2027.

Oct 10, 202614 min

GPT-5.4 Free API guide showing 5 access methods and pricing in 2026

GPT-5.4 Free API: Every Way to Access It and What Each Costs

As of October 10, 2026, GPT-5.4 has no free API tier. Paid orgs that share data with OpenAI can get up to 1M free tokens a day; otherwise it's $2.50/$15 per 1M.

Oct 10, 202620 min

Diagram headed Free tier is not free API calls: one API request is checked against usage tier, credit balance, and rate limits, and $100 a month is labeled a usage ceiling, not a credit grant.

OpenAI API Free Tier: What's Actually Free and What Still Bills

The Free tier's $100/month is a usage ceiling, not credit. Only a credit grant in your account or eligible data-sharing daily tokens make requests free.

Oct 9, 202613 min

Several Gemini API keys share one project's RPM, TPM, and RPD limits.

Gemini API Free Tier 2026: Limits, Free Models, and Free API Keys

The Gemini API key is free. As of October 8, 2026, Flash, Flash-Lite, 2.5 Pro and Gemma 4 are free; 3.1 Pro and image models aren't. Limits live in AI Studio.

Oct 8, 202619 min

Developer path from a controlled Google Cloud project to a protected Gemini API key and a successful first request

Google AI Studio API Key: Is It Free? Create, Secure, Verify

The Gemini API key from Google AI Studio is free. Flash, Flash-Lite and 2.5 Pro calls are free within project limits; image models and 3.1 Pro need billing.

Oct 8, 202610 min

GPT-6 Astra request workflow linking the service, project key, and model to response verification

GPT-6 Astra API Access: Check Your Project and Make a First Request

GPT-6 Astra API access depends on the service, organization, and project behind your key. Use a complete first-request example to check access, read the response correctly, and identify the next step when a call fails.

Oct 8, 20267 min

Conceptual Grok Voice API path from authenticated WebSocket input through response status and decoded WAV audio to speaker playback

Grok Voice Agent API: WebSocket Quickstart, Authentication, and Pricing

Use wss://api.x.ai/v1/realtime for a live Grok voice agent. Start with a server-side text or PCM turn, decode the audio reply, and use short-lived client secrets for browser connections. Current audio pricing is $0.08 per minute, with different duration meters for server VAD and push-to-talk.

Oct 8, 202612 min

Nano Banana Pro API quota by billing tier: about 74, 372 and 1,488 Pro images per 10 minutes on Tiers 1 to 3

Nano Banana Pro API Quota Increase: Tier 1–3 Caps in Images

Nano Banana Pro API quota grows with your billing tier, not with extra keys: Tier 1's $10 per 10 minutes buys about 74 Pro images, Tier 3's $200 about 1,488.

Oct 8, 202612 min

Nano Banana API docs and tutorial cover: GEMINI_API_KEY from Google AI Studio, four image model IDs, start with gemini-3.1-flash-image, and the October 2, 2026 legacy shutdown

Nano Banana API Docs and Tutorial: Key, Model IDs, and Code

Nano Banana is the Gemini API's image model family, used with an AI Studio key. Start new code on gemini-nano-banana-2.1; Nano Banana 2 is now deprecated.

Oct 8, 202621 min

Conceptual Nano Banana on Vertex AI cover showing the 2.1 model, enterprise client setting, global location and ADC or service-account authentication feeding a cloud image-generation workflow.

Vertex AI Nano Banana API: Model IDs, Setup, and Cost per Image

Start new Vertex AI projects on gemini-nano-banana-2.1 with enterprise=True and location global. Nano Banana 2 stays supported there until May 2027 or later.

Oct 8, 202621 min

An agent execution checkpoint blocks new calls, preserves validated work, and separates uncertain writes for reconciliation.

How to Stop an AI Agent Tool Loop—and Keep It Stopped

Stop new tool execution, preserve verified results, and reconcile any uncertain write before resuming. A turn limit contains a loop; progress checks and operation records make recovery safer.

Oct 7, 202612 min

Function-calling migration flow linking application execution and the returned result through the original call_id

Migrate Function Calling from Chat Completions to Responses API

Move the whole function-calling loop: validate the model’s arguments, run the authorized handler, return the result with its original call_id, and continue until the user gets a final answer. The JavaScript example preserves typed state and handles failed or unfinished responses.

Oct 7, 202614 min

Recorded transcription and live captions have different audio inputs, text states, and annotation options.

Gemini 3.5 Transcribe API: Recorded Audio, Live Captions, and a Safe First Build

Use Files and Interactions for recordings, or the Live API for raw PCM captions. Vocabulary biasing and recorded speaker or word annotations require separate configurations.

Oct 7, 202611 min

Agent, retry, helper, and fallback requests converge on one shared budget gate before reaching the model provider.

LLM Agent API Spend Kill Switch: Stop Runaway Costs Before the Provider Call

Block the next unaffordable model request before dispatch: reserve its maximum cost against the shared run budget, reconcile the billable result, and keep budget stops terminal across retries and helpers.

Oct 7, 202613 min

Three permitted storage policies feed an evidence-based approval for the exact API route and logging configuration.

LLM API Data Retention vs Zero Data Retention: A Route-Level Audit

Limited retention can fit an approved workload. ZDR requires evidence for the exact account, model, endpoint, features, and data path—and a gate that rejects unapproved changes.

Oct 7, 202611 min

Diagram connecting application JSON and YAML to an image API request, with a separate PDF document-processing step

Nano Banana Pro API Guide: JSON, YAML, Image Saving, and PDFs

Call gemini-3-pro-image through Google Interactions, save image bytes from the full response, and compile your own JSON or YAML job settings. PDFs need a document-capable model before Pro receives the verified facts and page images.

Oct 7, 202612 min

A failing worker tests its DNS, TLS, proxy, and endpoint while preserving ambiguous request outcomes.

OpenAI API Connection Error: Fix APIConnectionError by Testing the Route First

For APIConnectionError, capture the underlying exception and test DNS, TLS, proxy settings, and the exact endpoint from the failing runtime. An HTTP response narrows the problem; recovery still requires a successful request through the original application.

Oct 7, 202613 min

Three billing meters compare Realtime, an STT–LLM–TTS pipeline, and GPT-Live for the same completed task.

Realtime API vs Transcription Pipeline Cost: A Like-for-Like Budget

For transcripts, start with transcription. For a spoken assistant, compare Realtime—including mini—with the whole STT–LLM–TTS stack. A ten-minute call is not ten minutes of user audio plus ten minutes of assistant audio.

Oct 7, 202613 min

A decision map comparing a direct typed response with application-executed tools and a final typed object

Structured Outputs vs Function Calling: Choose the Right OpenAI API Contract

Use text.format when your application needs the model's final typed answer. Use function calling when the answer requires application data or an action. Combine them when a real tool result must become a schema-controlled object; then check both the schema and the business result.

Oct 7, 202611 min

Illustrated OpenAI API quota troubleshooting overview with billing, limit, project, and recovery checks.

OpenAI API Quota Exceeded Error: Fix insufficient_quota Without Chasing the Wrong Limit

For OpenAI API quota exceeded or insufficient_quota, inspect error.code before buying credits or retrying. A depleted balance, enforced spend cap, and approved usage limit need different fixes; ChatGPT subscriptions do not fund API calls.

Oct 6, 202610 min

Application configuration, conversation state, tool execution, and traffic controls in an Assistants to Responses migration

Migrate OpenAI Assistants API to Responses API: A Production Cutover Guide

The Assistants API has shut down. Restore configuration and available history from your own records, run custom functions through a complete Responses loop, and switch traffic with a recovery path that works today.

Oct 6, 202613 min