Docs
Change Log

Change Log

Documentation recording the GoAPI ongoing dev progress

20240724 | luma api update

  • added loop param for luma api.
  • Luma Dream Machine api now supports BYOA!

20240723 | suno api and luma api update


20240719 | gpt-4o-mini and uncensored llm

  • added gpt-4o-mini and uncensored-small-32k-20240717 models for LLM API, with pricing

20240716 | Captcha Resolver and Dream Machine start&end frame

  • Supported start & end frame of Dream Machine API, see luma api.
  • We added captcha checker, captcha webhook and automatic captcha resolver for our Midjourney BYOA users, you will see guides in byoa settings.
  • We published our doc page for all available notification webhooks.

20240622 | Claude models

  • Added Anthropic's models claude-3-opus-20240229 and claude-3-5-sonnet-20240620 with 75% Ahthropic's pricing, see: pricing and LLM API doc.
  • As our operation capacity grew, we are reducing midjourneyAPI pricing by 20% for turbo mode, details here: New MJ PPU Pricing

20240615 | LLM API massive update

  • Added reverse engineered model gpt-auto and gpt-4o-plus with amazing pricing.
  • gpt-auto is reverse engineered version of the Dynamic tab in ChatGPT: OpenAI determines when to use gpt-4o or gpt-3.5-turbo internally. In our test, most of the responses will be generated by gpt-4o.
  • gpt-3.5-turbo is fully backed by reverse engineered ChatGPT's gpt-3.5-turbo, therefore its pricing goes lower.
  • Due to the policy change of OpenAI, we are changing the pricing of dall-e-3, tts, whisper, embedding and original gpt-4o, please check the pricing.

20240528 | GPTs API and Suno API update


20240514 | GPT-4o update

  • Support gpt-4o in GPTS API
  • Support --sref random and --sv parameters in Midjourney API (V6 engine)

20240513 | Suno API update

  • Added Get Full Song API
  • Since May 1st, all USD top-ups will automatically generate invoices, you can download them in Top-up histroy in our dashboard's billing page.

20240418 | Suno API dropped

  • Added Suno API, and its pricing.
  • We are working around clock to provide BYOA and quota display for suno, stay tuned!

20240323 | Image Editing API and Dashboard Updates

Dashboard

  • Added Account status Data Report for MJ BYOA users. You can find it on MJ BYOA page in Dashboard, pricing.
  • Added Account Notifications for users on developer plan and beyond. You will see critical events on your account and automate them through webhook.

LLM API

  • Added gpt-3.5-turbo-0125 and gpt-4-0125-preview model, see detailed pricing.

MJ API

Added bot_id param

  • Enable user to decide which discord account handles what task. Available for BYOA user with Premium Plan. See pricing.
  • Only Imagine, Blend, Describe have this param. Check Imagine for API spec.

Image Editing API

We have gathered state-of-the-art models for popular image editing tasks!


20240312 | support --cref

Added

  • Character References are now supported, be sure to check the MJ official to learn how to use --cref!

Removed

  • Discord has officially changed the expire policy for its attachment cdn. Now discord attachment url serves for only 24 hours. Therefore, we removed the Image-Upload api.

20240218 | reduced pricing for V6 PPU

Added

  • V6 will be the default engine as per MJ changes, if you want to use V5.2 or other versions, you should put --v 5.2 in your prompt.
  • New MJ PPU Pricing Significantly reduced price for V6 PPU, Upscale and Describe. We will continue monitoring MJ's V6 GPU usage, and will change the pricing accordingly.

20240205 | improved faceswap

Added

  • Face Swap Async API is live! You can upload url or base64 for input, and get url or base64 in output for face swap. More accurate error messages and task status.

20240202 | Style References

Added

  • Style References are now supported, be sure to check the MJ official annoucements to learn how to use --sref! There's no official doc yet!

20240130 | Major Updates around V6 | GPTs BYOA

MJ

  • Niji 6 engine supported, use with --niji 6 inside your prompt. The Niji 6 PPU pricing goes same as V6 Pricing
  • Pan, Zoom and Vary(Region) supported for V6 tasks!
  • Repeated upscales and variations are now supported, which are equivalent to clicking the same mj button multiple times. (U1/U2/U3/U4 can't be repeated since its all same result)
  • For MJ PPU users only: We've updated a penalty schema to restrict Users that have mass relaxed job or invalid prompt inputs. Most PPU users will not be affected by this change. Details here

GPTs API

  • BYOA method of using GPTs API comes to live! Bind your ChatGPT Plus/Team account and chat with GPTs through API! Details here

LLM API

  • text-embedding-3-small and text-embedding-3-large are supported from our Embedding API

20240118 | Face Swap API and Image Urls

Added

  • Face Swap API is alive!
  • image_urls now available for some users in mj task result, details Here

20240117 | BYOA settings improvements

Added

  • A backup schema for BYOA account setting, see details HERE
  • Backup and Active(outside active period) status in BYOA account list, now user will have a more accurate overview of their BYOA accounts.

20240112 | Midjourney Special Upscaler

Added

  • Our Midjourney Upscale endpoint now supports:
    • light/beta upscaler for v4 jobs (Added just now)
    • 2x/4x upscaler for v5 jobs
    • subtle/creative upscaler for v6 jobs (Added just now)
  • Check MJ PPU Pricing If you are using PPU

20240111 | GPTs API can use all public GPTs

Added

  • GPTs store And ChatGPT Team Plan is officially launched by OpenAI, details HERE
  • gpt-4-gizmo is available in GPTs API, you can call any public GPTs out there, find spec HERE
  • Our GPTs API is compatible with both ChatGPT Plus and Team account.

20240109 | GPTs API(A Proxy for ChatGPT PLUS)

Added

  • Proudly introduce alpha test of GPTs API, in PPU mode. Check API spec HERE.
  • Working around clock to support upcoming GPTs store, before the end of this week.
  • Will provide BYOA for GPTs(ChatGPT PLUS) this month.

Changed

  • As written in email, we are closely monitoring the performance optimization of midjourney V6. Will reduce V6 PPU price if significant improvement was obeserved.

1221 | Support of Midjourney V6 Beta

Added

  • Add --v 6 in prompt to generate images with Midjourney's most powerful V6 engine on our Midjourney API !!
  • Midjourney has not yet change its default engine to V6, we will follow this decision. If you do not specify, all mj calls will be default to --v 5.2.
  • NOTE - V6 release will result in a huge wave of traffic to Midjourney. Expect Super Slow Speed for all relaxed generations.

1220 | Embedding, TTS and Whisper API

Added

Dashboard

  • We added a pause/start button in MJ BYOA account setting, user can now stop new mj task distributing to certain account.(Upscale/Variation/Zoom/Pan on old tasks will still be performed on that account)

1212 | Task Progress | MultiFetch

Added

  • Multi Fetch endpoint is now live, user can batch fetch up to 50 MJ tasks in one request.
  • task_grogress and intermediate_image_urls in MJ job processing are finally available for MJ's Fetch, MultiFetch and Webhooks. Result varies base on user's plan, check Optional Param and Pricing Plan

Dashboard

  • We added a weight selector in MJ BYOA account setting, user can now have more control over which account are more responsible for handling incoming request.(Low:Mid:High=1:2:4)

Changed

  • Discord has not yet change its CDN policy, therefore we temporarily paused tool/image_upload endpoint and auto_upload for MJ tasks. We will wait and see.

1128 | Image Upload API | Auto Upload

Added

  • Auto Upload for Midjourney Images: For users on Developer Plan or Above, we provide FREE auto upload for all Midjourney Images, that means there's an extra permanent_url in every mj task result (This decision is made upon: Discord News (opens in a new tab). Will provide a param for user to turn this off in the future).
  • Better domain support for LLM API non-stream mode: we provide a proxy domain to better handle non-stream call to GPT completion apis. Details: LLM API
  • Task History: we rolled out our task history page on our dashboard(Only MJ and SD tasks are visible for now).

1115 | GPT-VISION | DALL·E 3

Added

  • gpt-4-vision-preview now live for users on Developer Plan or above. You may notice that this model returns slightly different data structure in completion API: Vision model change
  • The new Dall·e 3 endpoint went live! DALL·E 3 API

Referral Program

  • You can check the referral page in dashboard for more detail:Referral

1113 | Midjourney Upscaler | New GPT models

Added

  • Midjourney Blend API now support optional dimension. See BLEND API
  • Midjourney Upscale API now support 2x and 4x index. API spec: UPSCALE API Pricing: MJ PPU Pricing
  • LLM API now support GPT-3.5-turbo-1106 and GPT-4-1106-preview model with Even cheaper price. API spec LLM API Pricing: LLM Pricing

Referral Program

  • We are rolling out our referral program, you can now invite your friend and gain credit grants! More details will be talked about in our discord channel, stay tuned!

1101 | Dashboard | New Pricing

Added

  • We released dashboard for GoAPI, Dashboard, the user manual: Dashboard Mannual
  • We have applied the new pricing plan, for more details please check Pricing Plan
  • Relax and Turbo mode are live (For Midjourney API)
    • We now allow --fast, --relax, --turbo parameters inside midjourney prompt
    • For old Mixed mode, it will be default to Relax, but if the prompt comes with --fast or --turbo, it will perform in that mode.
    • We calculate the cost by the actual execution mode of the job.

Changed Small changes on the midjourney task result, it's still backward compatible, as below

  • Midjourny Task Data Structure Changelog:
    • Meta
      • added quotaFrozen for new MJ quota usage
      • frozenCredit will always return 0
    • Result
      • added quotaUsed for new MJ quota usage
      • credit will always return 0
      • FOR BYOA USERS, discord_image_url will return the discord attachment url

Removed

  • Removed all the payment button on our homepage. All payments are integrated into our Dashboard.

1018 | Midjourney API | Seed Endpoint

Added

  • SEED API endpoint now availiable: User can make a seed request with task_id after /imagine, then when you fetch result, can get seed value in task_result.

0928 | LLM AP | New Primary API domain

Added

  • We aggregated OpenAI's ChatGPT api into GoAPI. Since we are alpha testing this feature, the price was labeled 80% CHEAPER than OpenAI(For GPT-3.5), you might want to check it out LLM API We enabled new primary domain for all api endpoints: api.goapi.xyz, this domain will be the recommended domain for all new user and new api calls. As for api.midjourneyapi.xyz, it's still fully functional and will be maintained as our backup domain. So there's no need to modify your existing code or app.

Deprecated

  • We retired the V1 version of Midjourney API in Midjourney API V1, all request to V1 endpoints will get an error. We will reroute V1 endpoints to V2 on November, 2023.

Announcement

  • We are planning to change our pricing plan by November 1st. Many infomation rollouts upcoming in October. Stay tuned for our updates & annoucements in our Discord and other channels.

0912 | Action Checker

Added

  • Implemented the recording of actions in an Action Tree, all actions to be performed only once, except for the reroll action. See Action Checker

0912 | Prompt Checker | Reroll & Inpaint

Added

  • Validation endpoint now available: Validation Users can verify their prompts against our predefined checks without triggering the imagine request
  • Reroll endpoint now available: Reroll adds flexibility to the user experience by allowing you the ability to initiate re-drawings based on prompts and the parent task
  • Inpaint endpoint now available: Inpaint allows for partial modification of the image in the selected region. The unique aspect of this feature is that it relies on the base64 encoding of the designated mask, providing a more precise editing tool

Enhancements

  • skip_prompt_check: For users who prefer a streamlined workflow, we've added the skip_prompt_check option. This allows bypassing our internal prompt validation process, offering more direct access to task execution.

0906 | Accout Management API | Redeem code

Added

  • Redeem Code endpoint now available: Users can redeem credits to their existing account. Please refer to our Dashboard to redeem credits.

0828 | Midjourney API | Discord Image URL

Added

  • discord_image_url: A new attribute associated with the task_result returned by the Midjourney API V2 fetch. This attribute contains the Discord original URL of the images, granting users direct access to the original source of the image on Discord.

0822 | Midjourney API | Blend & Describe

Added

  • Blend and Describe Endpoints Now Available: The /v2/blend and /v2/describe endpoints are now fully operational and available for utilization. These offer enhanced features and improved performance over their previous iterations.

0817 | Midjourney API | Pan & Outpaint

Added

  • Midjourney API V2: We're excited to announce the release of our Midjourney API V2! This new version includes Pan and Outpaint, and multiple new params such as 'process_mode'. Check out the Midjourney V2 API for more details.

Deprecated

  • Midjourney API V1: Please be aware that as of this release, API version 1 is now deprecated. While the V1 will continue to function for the October 1st, 2023, no further development or updates will be made. We encourage all users to migrate to the new API V2 and take advantage of the improved features and performance.

Midjourney API v1 will officially be retired on October 1st, 2023. After this date, any requests made to v1 will fail.


0807 | Stable Diffusion API | Inference steps & guidance scale

Added

Redeem a code