Changelog

Join our Discord Server or Telegram Channel to stay in touch.

December 13, 2024

  • The Mureka API v1 now supports all functionality of Mureka AI, including Pro subscription features. We have added the ability to regenerate and extend music, upload custom soundtracks for reference, and use melodies (motifs) as references.

    Examples.

  • Added parameter static to Runway Gen-3 Alpha Turbo POST /runwayml/gen3turbo/create

December 10, 2024

We released support for the Runway Gen-3 Alpha Turbo Act-One gen3turbo/actone and an update for Gen-3 Alpha Act-One gen3/actone. Now you can use video assets for your character, specify the level of expressiveness in motion and choose between portrait or landscape mode for image characters.

Example • Gen-3 Alpha Act-One

The final video-to-video result:

The final image-to-video result:

Driving video:

Character image:

Character video:

December 6, 2024

We’re proud to announce experimental PixVerse API v2 for the PixVerse.ai. PixVerse.ai generates videos from text and image prompts, supports a large number of video effects, and can extend, upscale, and lipsync videos.

Example • PixVerse Effects and Lipsync

The final video features a 32-second clip created with four different effects and lip-synced with the uploaded soundtrack.

Song composed using Mureka API v1.

Original image

December 3, 2024

The MiniMax POST videos/create has been updated to support the recently introduced I2V-01-live model. You can now optionally specify the desired model with one of the following values:

  • T2V-01 (default) model for text-to-video
  • I2V-01 (default) model for image-to-video with optional prompt
  • I2V-01-live model for image-to-video with optional prompt, added December 3, 2024

December 2, 2024

We’re proud to announce experimental Mureka API v1 which provides full access to Mureka AI, a music generation service from Chinese tech giant Kunlun Tech featuring proprietary SkyMusic 2.0 model and positioned to compete with Suno and Udio. Mureka AI creates unique songs based on user-provided lyrics or description (AI-generated lyrics), selected musical styles, and references to specific vocals and songs, including those uploaded by the user.

See Mureka API Music Samples for more details.

November 24, 2024

Added support for Expand Video (outpaint) with Gen-3 Alpha Turbo POST gen3turbo/expand.

Example • Gen-3 Alpha Turbo Expand Video

Prompt:

Dancing at a disco club

Expanded video (landscape)

Original video (portrait)

November 22, 2024

Added support for Midjourney inpainting Vary Region and Vary Region + Remix. The POST jobs/button endpoint parameter mask accepts a base64-encoded black and white mask image. Feel free to use the Vary Region mask editor.

Example • Midjourney Vary Region

Original POST jobs/imagine prompt:

Princess Peach poses as a criminal boss, smiling and striking a pose in front of a Mario-themed Bentley go-cart. –ar 9:16

POST jobs/button button Vary (Region) with the remix prompt and a mask:

Princess Peach poses as a criminal boss, smiling and striking a pose in front of a Mario-themed Bentley go-cart. She is wearing a Mario hat.

November 18, 2024

  • Updated Gen-3 Alpha Video to Video gen3/video and Gen-3 Alpha Turbo Video to Video gen3turbo/video to support video generation longer than 10 seconds (see examples below).
Example • Gen-3 Alpha Turbo Video to Video (16 sec long with music)

Prompt:

A pretty lady with blonde hair, wearing a swimsuit, is dancing in space. The Milky Way is barely visible in the background, while a brilliant spotlight beams down from above, illuminating her like a star on a stage in the movies.

structure_transformation: 9

Final 16 seconds result

Original 16 seconds video

Example • Gen-3 Alpha Video to Video

Prompt:

A pretty girl wearing a swimsuit dancing gracefully on a tropical beach.

structure_transformation: 9

Final result

Original video

  • FaceSwap swapid and picsi endpoints will return the response code 412 when the account(s) run out of credits.

November 8, 2024

  • Added support for Super-Slow Motion. Super-Slow Motion uses the power of AI to generate and add new frames to your clips for a crisper, cleaner slow-motion effect on any video of your choosing. This endpoint offers free & unlimited Super-Slow Motion generations. It is available to users with free accounts without any limitations. You can run as many parallel jobs as you want. Additionally, this endpoint can be used to export your mp4 videos to Apple’s ProRes 4444 format.
Example • Super-Slow Motion

Original 10 seconds video

Final 20 seconds result using speed 0.5 (2 times slower)

November 5, 2024

November 4, 2024

  • We released support for Gen-3 Alpha Turbo Video to Video gen3turbo/video.
Example • Gen-3 Alpha Turbo Video to Video

Prompt:

An ultra-realistic underwater music video features the enchanting Princess Ariel, who is wearing a tiny swimsuit. She dances gracefully amidst a vibrant scene with fish and sharks gliding around her. The water is alive with the magical glow of sunlight streaming through from the surface, creating a mesmerizing array of bubbles rising all around.

structure_transformation: 9

Original video

Final result

  • Gen-3 Alpha Turbo Text/Image to Video gen3turbo/create updated to include camera control (horizontal, vertical, roll, zoom, pan, tilt)
Example • Gen-3 Alpha Turbo Image/Text to Video

Prompt #1:

Smiling and winking.

zoom: 2 vertical: 2 pan: -10

Original image

Final result

Prompt #2:

The camera flies right into the dragon’s mouth, filled with fire.

zoom: 10

Original image

Final result

  • Gen-3 Alpha Video to Video gen3/video updated to include seconds parameter.

  • The article Fun with MiniMax API has been updated to deal with 502 and 504 MiniMax response codes (MiniMax website is too busy).

  • Recommendations for maxJobs updated for MiniMax accounts/post and videos/create:

    Please set maxJobs number according to your subscription plan. We recommend setting it to 2 (3 maximum) for Free, and 3 (4 maximum) for Standard and Unlimited plans. Although you can set it to higher values (3 for Free and 5 for Standard and Unlimited respectively), it won’t make any difference since only one video can be processed at a time for Free accounts, and no more than two for Standard and Unlimited accounts. It’s helpful to have a spare empty slot for cases where the MiniMax website is too busy and responds with 502 or 504 while still accepting jobs into the internal queue.

October 28, 2024

We released support for amazing Runway Act-One gen3/actone.

Example • Gen-3 Alpha Act-One

Character image

Driving video

Final result

October 25, 2024

October 23, 2024

  • Response code 596 handling added videos/create to deal with temporary account holds by hailuoai.video.
  • The article Fun with MiniMax API script updated to handle response code 596.

October 18, 2024

  • Article Create Runway videos like a Pro has been updated to use a NodeJS script instead of Bash, and now supports all Gen-3 Alpha and Gen-3 Alpha Turbo options.
  • Response code 412 has been added to all applicable Runway API endpoints to indicate an insufficient credits response.

October 15, 2024

  • The article Fun with MiniMax API has been updated to support image prompts and now features more generation samples.
  • A new GET scheduler/available endpoint has been added to retrieve the list of videos currently running via the API, along with the available account capacity.
  • Support for multiple new response codes has been added for POST videos/create, including 412 for Insufficient credits and more.
  • Q&A MiniMax section updated with execution time cost estimations.

October 14, 2024

We have introduced the capability to upload images using POST /files and query images via GET /files. This allows you to specify images while utilizing the videos/create endpoint. Additionally, you can use GET /features to access your hailuoai.video account information.

October 11, 2024

Experimental MiniMax API has been updated to work with hailuoai.video. You will need a Google account to log in to hailuoai.video. Our API supports both free and paid hailuoai.video accounts. The article Fun with MiniMax API has also been updated for your convenience.
Ability to add image in addition to text prompt will be released Monday October 14. 2024.

October 10, 2024

We created designated Gen-3 Alpha Turbo endpoint and added support for first and last frames.

Original Gen-3 Alpha endpoint will no longer accept turbo and aspect_ratio params starting November 1, 2024. Please consider switching all turbo load to Gen-3 Alpha Turbo before deadline.

October 7, 2024

We released jobs/seed_async, which will fully replace the deprecated jobs/seed starting November 1, 2024. The newly added jobs/seed_async has a 100% success rate, whereas the now deprecated jobs/seed was failing to retrieve seeds under some obscure conditions. Please consider switching to use the new version before November 1, 2024.

October 4, 2024

We added music generation to experimental MiniMax API v1.
Check out MiniMax API Music Samples.

September 30, 2024

September 25, 2024

We’re proud to announce experimental MiniMax API v1.

MiniMax AI, a Chinese AI startup backed by Alibaba and Tencent. MiniMax unveiled its latest text-to-video AI model at Hailuo AI in early September 2024. This model is notable for generating ultra-realistic video clips with precise prompt following.

September 24, 2024

  • Added support for Gen-3 Alpha Turbo parameter aspect_ratio. Supported values: landscape (default), portrait.

September 14, 2024

August 31, 2024

August 23, 2024

August 19, 2024

August 16, 2024

  • Added support for turbo mode for Gen-3 Alpha.
    Turbo mode provides significantly faster generation times with 50% cost reduction, at 5 credits per second of video generation. Turbo mode is supported with free accounts and is the only way to experience Runway Gen-3 Alpha if you have a free account.

August 12, 2024

August 8, 2024

We’re proud to announce experimental Runway API v1.

July 23, 2024

Maintenance update for experimental InsightFaceSwap API v1.

  • Hotfix released to align with InsightFaceSwap bot changes released July 22, 2024
  • faceswap/swap is deprecated, please use faceswap/picsi instead.

June 12, 2024

Major update for experimental InsightFaceSwap API v1.
Following new API endpoints were added:

June 10, 2024

Create breath-taking videos with PixVerse AI.
PixVerse API demo

June 3, 2024

Please welcome our new addition, an experimental API for the PixVerse Discord Bot by PixVerse.AI. PixVerse currently supports text and image inputs for generating 4-second-long videos. The available video styles include Realistic, Anime, and 3D Animation. For the Anime style you can reference one of the anime characters. PixVerse can also create a meme face video from a provided image.

February 19, 2024

Article Face swap and animate images generated by Midjourney using InsightFaceSwap and Pika (using webhook) includes both JavaScript and Python code, along with generated examples.

February 12, 2024

We’re proud to announce experimental REST API for InsightFaceSwap Discord Bot by Picsi.Ai. InsightFaceSwap allows users to swap faces from source image(s) onto different target images. It offers free and paid subscription models. Paid subscribers have access to a wide selection of extra features such as HiFidelity Mode, ARTIFY, oldify/youngify, morphing multiple faces in one image, and many more.

January 26, 2024

Added jobs/seed endpoint to retrieve the seed for for the following completed jobs:

January 22, 2024

Added support for Remix Mode, when using jobs/button endpoint the buttons V1, V2, V3, V4, Vary (Strong), Vary (Subtle), ⬅️, ➡️, ⬆️, ⬇️ and 🔄 now accept a prompt parameter. You can toggle /prefer remix via jobs/remix.

January 15, 2024

We’re proud to announce experimental REST API for Pika Discord Bot by Pika.art. Pika currently support text and images as input for video generation, and private generations by directly messaging the bot. The Pika API does not require any additional subscription, and we believe it will complement our current Midjourney API offerings well.

January 5, 2024

  • We’ve added the ability to use multiple Midjourney accounts, complete with automated load balancing. See How to use multiple Midjourney accounts for more details.
  • The format for jobid’s has changed to include the channel: prefix. Previously, it was job:20240106053233367-user:791 now it is job:20240106053233367-user:791-channel:1170809721663422494.
  • Response codes 412 and 413 have been eliminated, we now use response code 400 instead.
  • If you are already using v2, no changes are required. To switch from v1 to v2, simply update the API URLs from …/v1/… to …/v2/….

November 28, 2023

November 22, 2023

November 18, 2023

API v2 released

November 13, 2023

November 3, 2023

October 30, 2023

October 16, 2023

October 09, 2023

New functionality added:

Updated:

  • Website documentation and samples.
  • Postman collection.