AI image, video, and audio generation with persistent artifacts.Ship with one API.
Call leading generative models, track each task or wait for callbacks, and receive persistent artifact links without extra storage work.
Sign up for 200 Credits
Available through one API
Featured image, video, and audio models
Start with curated models for image generation, editing, music, and video workflows.
One task API for every generation workflow
Submit a job, check status, receive a callback, and retrieve stored artifacts through the same simple lifecycle.
modelPass the base model name to call any model through the same endpoint.
requiredinputModel-specific parameters, validated per request.
requiredrouteOptional route such as beta or ext; omit it or pass default for the default route.
optionalstorageSet persistent to get low-cost durable artifact links back.
optionalcallbackGet notified on completion instead of polling.
optional
curl https://api.hiapi.ai/v1/tasks \
-H "Authorization: Bearer $HIAPI_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: order-42-submit" \
-d '{
"model": "gpt-image-2.5-flare/text-to-image",
"route": "default",
"input": {
"prompt": "a glass speaker, studio light",
"aspect_ratio": "1:1",
"resolution": "1K"
},
"storage": "persistent",
"callback": {
"url": "https://example.com/cb",
"when": "final"
}
}'Ready for developers and AI agents
Integrate through API docs, MCP, Skills, and llms.txt so developers and agents can discover, call, and compose HiAPI.
API docs
Reference material for direct developer integration.
MCP
Agent-accessible tools for HiAPI workflows.
Skills
Reusable agent instructions for common workflows.
Use HiAPI in any agent
Provide the llms.txt prompt to your agent so it can understand HiAPI models and APIs, then select Skills, Remote MCP, or direct API access.
Read this HiAPI documentation index and answer my questions based on it: https://www.hiapi.ai/llms.txt About HiAPI: a unified AI API gateway for image, video, audio, and text models, with Agent Skills, Remote MCP tools, text endpoints, and a unified async API. My question: I want to generate images, videos, audio, or text with HiAPI. First decide whether Skills, Remote MCP, or the API fits best, ask one clarifying question if necessary, then provide the shortest setup procedure.
Know the cost before you ship
Every model lists its price up front, so you can size usage before you integrate.
Transparent pricing
Review each model's price before integration.
Predictable cost
Estimate cost clearly before a production launch.
Cost-aware routing
Use the route field to select an available route such as default, beta, or ext and compare costs clearly.
Volume discounts
Larger usage can get volume-based rates. Contact sales
Built for production generation workloads
Reliable access, persistent outputs, task tracking, and support for long-running image, video, and audio jobs.
Availability-first routing
Designed for stable access, not only the cheapest available route.
Persistent outputs
Generated artifacts stay reachable through returned links.
Optimized for Long-running jobs
Task tracking and callbacks fit image, video, and audio workloads.
24/7 support
Coverage for production integration and operation.
Feedback from real generation workflows
Representative scenarios from teams building AI video tools, creative automation, and agent workflows.
We test video models quickly, keep generated assets reachable, and skip building a storage path for every provider.
Persistent artifact links make it easy to pass outputs between campaign tools, review, and downstream automation.
A task API with callbacks gives agents a clean way to request, wait, retrieve, and reuse generated media.
Frequently Asked Questions
Quick answers on supported models, persistent artifacts, callbacks, and trying before you build.
Which models does HiAPI support?
HiAPI supports selected image, video, and audio generation models, including GPT Image 2.5 Flare and Sunburst. Browse the full list on the models page.
How do persistent artifact links work?
Set storage to persistent and generated outputs are stored for you and returned as durable links in the task result.
Can I use callbacks instead of polling?
Yes. Provide a callback URL and HiAPI notifies you when the task reaches its final state, so you skip polling.
Can I try models before integrating?
Yes. Use the Playground on each model page to explore parameters and example outputs before writing code.
Start building with persistent AI artifacts
Get your API key and call image, video and audio models through one task-based API.
Guides & resources
Tutorials, model reviews, and pricing breakdowns to help you ship faster.
Getting started
hiapi Image and Video Generation APIs: The Complete Platform Guide
hiapi is an image and video generation API gateway: one Bearer-auth endpoint fronts FLUX 1.1 Pro, nano-banana, GPT-Image-2, Qwen-Image, Wan2.7, Seedance, and HappyHorse. Pricing is per finished task, not per token — $0.02 for gpt-image-2 beta, $0.05 for nano-banana or flux-1.1-pro, $0.729 per second for 1080p Seedance footage. No token billing, no surge. All generation runs through the async POST /v1/tasks endpoint. Same shape for text-to-image, image-to-image, text-to-video, image-to-video. Most image models accept input.prompt + input.aspect_ratio. qwen-image-2.0 uses a DashScope-style nested schema with literal pixel sizes — keep a model→payload adapter map. Four sample images in this post — cover illustration, photoreal portrait, CJK calligraphy signboard, layout poster — have a current estimated budget through hiapi of $0.172.
GPT Image API Guide: Model IDs, Routes, and Migration
OpenAI's current direct API model ID is gpt-image-2; gpt-image-1 and gpt-image-1.5 are previous, deprecated image models. New HiAPI integrations should use gpt-image-2/text-to-image or gpt-image-2/image-to-image through POST /v1/tasks. ChatGPT image generation, OpenAI's Image and Responses APIs, and HiAPI's async task API are different product surfaces with different request contracts. The retired HiAPI Pro IDs map to the current text-to-image and image-to-image IDs; do not preserve Pro in new code. This page is the navigation hub. Detailed parameters stay in the two model docs, and current rates stay in the pricing guide and live pricing page.
Reviews & comparisons
GPT Image 2 vs 2.5: Image Quality, Editing, and When to Upgrade
Image 2.5 focuses on reference fidelity, targeted edits, and consistency across revisions. Flare suits fast creative work; Sunburst prioritizes precise editing. Choose by task and keep Image 2 for workflows that already produce reliable results.
GPT Image 2 Review: A Hands-On Honest Take
Text rendering is the real upgrade — English headlines, CJK calligraphy, in-image labels and prices all render correctly the first time. Photorealism is genuinely strong for product photography and detailed interiors. Portraits are good but occasionally read as slightly 'too symmetric'. Speed is a step backward from GPT Image 1.5: ~107 seconds per image versus 18–36 seconds. Plan workflows around it. Hard limit: no transparent backgrounds. If you need PNG cutouts, you're still going to need a second model or a background remover.
Pricing & cost
GPT Image 2 API Pricing: Cost per Image vs OpenAI, fal and Kie
HiAPI matches Kie at 1K, costs less at 2K and 4K, and sits below the medium-quality price on OpenAI, fal and WaveSpeed. On HiAPI, GPT Image 2 is $0.03 / $0.04 / $0.06 at 1K / 2K / 4K, for text-to-image and image-to-image alike. Real outputs at all three resolutions show when 2K or 4K is worth paying for.
Free AI Image Generation API, No Key Required: We Tested 4 in 2026
Pollinations' old URL returned one watermarked image, then HTTP 402 asking $0.01 in USDC per image. AI Horde's anonymous key 0000000000 works at up to 580×580; one of four harmless prompts was blocked as NSFW. With a key, GPT Image 2 on HiAPI costs $0.007 per image and followed the same prompts far more closely.