AIFAST LINK START

Get started with AIFast Link

Connect to GPT, Claude and Gemini APIs. Choose your next step: make a first call, verify an existing endpoint, migrate a project, or discuss enterprise usage and invoicing.

After registration, create an independent Key and confirm the model ID and price from the real-time catalog; then migrate the production business after completing the minimum request.

Link GitHub setup examples (Chinese / English): Protocol setup, troubleshooting checklists and downloadable offline cost examples.

GPT: Confirm the model ID in the live console
Claude: Verify through the Anthropic Messages endpoint
Gemini: Separate group; choose native or compatible endpoints by model
Corporate invoicing: Contact customer service for specific plans to confirm

Choose your task

Which step needs to be solved most now?

First time useComplete the first API request

Open 3-minute tutorial Existing APIDetect API gateway compatibility and routing risks

Check model declarations, tokens, dynamic questions, SSE and tool calls.

Run a free test Project migrationAccess OpenAI Compatible API

Keep the business parameters and replace the address, Key and model ID item by item.

View migration steps Business needsConsultation on high usage, recharge and invoicing

Prepare monthly usage, model, concurrency and invoice requirements and contact us for confirmation.

Contact Enterprise Support

Compare usage costs first

Pay-as-you-go or an API monthly plan

Link offers pay-as-you-go funding and GPT / Claude API monthly plans. Compare the amount paid for the same tasks before choosing a billing method. A plan price does not imply a fixed token price, and you should not assume it covers Gemini groups.

AIFast Credit is a platform usage unit, not cash. Estimate a task using model prices, then compare a monthly plan using your active days and daily peaks. Check the current plan terms before payment.

Check funding and monthly-plan rules

Estimate task costs and daily overages

View current model prices

Checklist before choosing a billing method
Option	What to check
Pay as you go	Amount paid, credited allowance, model and group prices
GPT / Claude monthly plans	Eligible groups, daily limits, validity, overages and wallet fallback
Gemini	Check its separate group; do not assume GPT / Claude plans cover it

Use the credited allowance shown on the funding page and the final payment and plan scope shown at checkout. An API monthly plan is not a model vendor's consumer subscription.

Shortest verification path

Registration is not the end point, access is completed only if the request is successful.

Don't change the entire project at once. Confirm authentication, model ID, and protocol with a minimal, reproducible request, and then gradually add business capabilities.

View the production go-live checklist

Verification 01 · Query model directory

curl https://www.aifast.link/v1/models \
  -H "Authorization: Bearer $AIFAST_API_KEY"

Expected result: HTTP 200, and returns the model ID callable by the current Key.

01
Confirm model and price
First check whether the target model is currently open, the API type, and the billing caliber.
02
Create independent API Key
Use independent keys for different projects and environments to facilitate quotas, troubleshooting, and cancellation.
03
Query the real model ID
Request /v1/models, do not write the web page display name directly into the code.
04
Send minimal request
Validate short text first, then test SSE, tool calls, image input, and error handling.

Check before making a decision

Price, compatibility, stability and exit methods should be looked at together

Does the ability match?

A successful Chat Completions request does not prove that Responses, Anthropic Messages, Gemini native endpoints, tools or multimodal features all work.

View compatibility matrix

Is the evidence reviewable?

Use status codes, response models, tokens, business question sets, and multi-period samples to leave reviewable results.

Open comparison template

Are the platform facts consistent?

The dynamic model, price and open status are subject to the console and real-time API, and page copywriting is not regarded as a permanent commitment.

View platform facts

High-intent access scenarios

Continue from specific tooling, model capabilities, or procurement needs

No need to read the entire document first. Choose the scenario closest to your current project, check the functional boundaries first, and then enter model pricing, registration or enterprise support.

development tools

Cursor and Claude Code

First confirm the custom Base URL entry, current version function boundaries, real model ID, and whether the OpenAI Key conflicts with the third-party Key.

Knowledge base and workflow

Dify chat models

AIFast Link provides GPT, Claude and Gemini chat models. Knowledge-base embeddings and reranking require separate services; successful chat does not mean the full knowledge base is configured.

Model scope

GPT, Claude and Gemini models

AIFast Link currently offers GPT, Claude and Gemini. Confirm the model ID, protocol endpoint and pricing in the live catalog before creating a minimal test.

Corporate Procurement

Unify purchasing, billing and invoicing

Prepare main models, estimated monthly usage, peak concurrency, project isolation, bill verification and invoice requirements before contacting, making communication more efficient.

Enter by decision stage

There is no need to register directly yet, and the next steps are clear.

information visitorTest existing APIs for free firstCheck model declarations, tokens, dynamic questions, SSE and tool calls.Run the test comparison visitorSee current models and pricesCheck the model ID, capability type, open status and real-time billing caliber.Check price Ready to accessRegister and create a test keyComplete the model list and minimum requests first before migrating the official project.Start registration corporate visitorsSubmit usage and invoice requirementsPrepare model, monthly usage, concurrency, project and invoicing information and contact us for confirmation.Enterprise support

Public caliber:Link offers GPT, Claude and Gemini models. Check the console, pricing page and real requests for current model IDs, prices, maintenance status and capabilities. Verify OpenAI-compatible, Responses API, Anthropic Messages and Gemini native endpoints separately; one chat request is not a full compatibility test. Gemini uses a separate group and is not included in GPT or Claude monthly plans.

Frequently Asked Questions Before You Get Started

Make the access boundaries and judgment basis clear before deciding on the next step.

What should you do first when using AIFast for the first time?

First check the model and price to confirm that the required model is currently open; create an independent API Key after registration, query the real model ID through /v1/models, and then send a minimum text request.

How are ⭐️ AIFast Credits settled?

AIFast Link payment currency, funding amount and credited balance follow the live checkout. Check current payment methods, discounts and the final charge before paying.

How to migrate existing OpenAI Compatible projects?

First replace only the Base URL and API Key, keep the model, prompt words and parameters unchanged to complete the minimum request; after successful verification, then test the streaming output, tool call, image input and timeout retry respectively.

How should I choose between pay-as-you-go and a monthly plan?

Compare the amount paid for the same tasks. Estimate pay-as-you-go costs for irregular usage; before choosing a monthly plan, check eligible models and groups, daily limits, validity and overage billing. A monthly plan is not necessarily cheaper and is not a ChatGPT or Claude consumer subscription.

How to determine whether there are compatibility or routing issues with the existing gateway API?

Use the temporary limit Key to run model detection, and combine the response model, token, dynamic questions, SSE, tool calls and other sub-evidence judgments. The test is a risk screening, not a model manufacturer certification.

What should enterprise teams do if they need invoicing or higher usage?

First sort out the estimated monthly usage, main models, concurrent requirements and invoicing information, and then contact customer service to confirm the current enterprise access, recharge and invoice plan; the final scope is subject to customer service and console information.

Can image, video, embedding, and reranking models use the same endpoint as text models?

Link offers GPT, Claude and Gemini models. Chat, image input and image generation are different capabilities to verify for each model and protocol. The current catalog lists no standalone video, embedding or reranking model. Configure separate embedding and reranking services for knowledge bases.

Where can I confirm the current real model ID?

After registering and creating an API Key, please query the model IDs available in the current account through the console or GET /v1/models. Model names in articles and pages are for illustration only and are not a substitute for live return values.

When you’re ready, start with verifiable first steps

First check the model and price, then register and create a Key; if the enterprise needs to contact customer service first to confirm the current plan.