Changelog
Console announcements
Section titled “Console announcements”Synced from the main console, newest first.
2026-11-10 18:26 (UTC+8) · #26
Section titled “2026-11-10 18:26 (UTC+8) · #26”🚀 Omni 1.2 Live
⏱️ Up to 60s
🎬 Gemini Omni 1.2 Video Model Officially Launched!
🎯 Core Model: omni-1.2
Full support: t2v | i2v | r2v | v2v, supports up to 60s ultra-long video generation!
✨ Enhanced Reference Consistency & Natural Lip Sync
Stable facial details across camera movements; supports precise multi-speaker alignment and voiceover scenes.
🎥 Reduced Frame Drift & Cinematic Smooth Camera Movement
More controllable actions and transitions; smooth pan, tilt, and zoom shots with strong visual impact.
🔍 Visit [Model Square] for detailed pricing and parameter options
Async task API calls, native audio/video lip sync, billed on success. See Square for details.
🚀 Try Omni 1.2 in [Model Square] Now!
2026-11-06 14:09 (UTC+8) · #25
Section titled “2026-11-06 14:09 (UTC+8) · #25”🚀 GPT Image 2.6
⚡ 50% Speed Boost
🎉 GPT Image 2.6 Is Officially Live!
⚡ Ultra-Fast Generation & Dual Architecture
Generation speed boosted by 50%, continuing dual nodes: Flare (Fast) & Sunburst (High-res).
🎯 Markup Inpainting & Native GIF Export
Added Markup Doodling for precise local edits, with direct support for GIF animations.
🧠 Complex Scene Reasoning & Planning
Supports pre-render thinking depth adjustment for enhanced composition, typography, and consistency.
🔍 Visit [Model Square] for Pricing & Parameter Details
Includes Flare and Sunburst endpoint mappings and billing rules in the console Model Square.
🚀 Go to [Model Square] & Try GPT Image 2.6 Now!
2026-11-06 11:45 (UTC+8) · #24
Section titled “2026-11-06 11:45 (UTC+8) · #24”🍌 Banana 40% Off
🚀 Gemini 4 Argon Live
📢 Service Updates & Model Overview
🍌 Banana Image Series Special Promotion Live
Includes Nano Banana 2.1, up to 40% off, as low as 6 fen (¥0.06) per image!
🚀 New Flagship Model Live: Gemini 4 Argon
The flagship model with a new architecture and enhanced performance is now fully deployed for API access.
🎬 Grok Video Model Price Adjustment
Due to rising costs from upstream system revisions, video generation pricing has been slightly adjusted.
🔍 Check [Model Square] for Detailed Specs & Pricing
Visit the Model Square in the console for exact pricing rules, available variants, and parameter specifications.
🚀 Visit [Model Square] Now to Explore & Test!
2026-10-28 23:06 (UTC+8) · #23
Section titled “2026-10-28 23:06 (UTC+8) · #23”🚀 Service Restored
🛡️ Small Trial Advised
🎬 Grok Video Generation Service Restored
Official rules and channels have been adjusted; Grok video generation is now back online!
Restored Model: Grok Video (Video Gen)
Channels have been re-established; users can now initiate video generation calls as normal.
💡 Usage Safety Tips:
Testing Recommended: Users are advised to first conduct
2026-10-18 22:18 (UTC+8) · #22
Section titled “2026-10-18 22:18 (UTC+8) · #22”⚠️ Official Policy Update
🔄 Recommended Alternatives
🎬 Notice: Grok Video Generation Service Suspended
Due to upstream Grok plan changes, Grok video generation is temporarily suspended.
Suspended Model: Grok Video
Resumption time will be announced later. Please watch platform announcements.
✨ Recommended Seamless Alternatives (Stable & Ready):
Omni 1.1 — High-quality images and smooth video generation
MiniMax H3 — Excellent dynamic motion and precise prompt adherence
Please update and switch model identifiers in your API code promptly.
🙏 Thank you for your understanding. We are closely monitoring progress!
2026-10-09 01:39 (UTC+8) · #21
Section titled “2026-10-09 01:39 (UTC+8) · #21”🚀 Major Model Update
🔥 Up to 40% Off
🍌 Nano Banana 2.1 Launch & Special Sale!
Next-gen Banana image generation model deployed; prices slashed across the series!
Banana Image Series Special Discount: Up to 40% Off
All-new Nano Banana 2.1 as low as 6 fen / image!
✨ Nano Banana 2.1 Key Upgrades:
Enhanced Prompt Adherence: Accurately follows complex prompts and nuanced details
Lighting & Texture Optimization: Significantly enhances realism in portraits, materials, and lighting
Ultimate Cost-Efficiency: Only 6 fen per image, combining top quality with low cost
2026-10-07 16:09 (UTC+8) · #20
Section titled “2026-10-07 16:09 (UTC+8) · #20”🇨🇳 National Day Deal
🔥 Limited 40% OFF
🎉 Happy National Day! Image Special Sale Is Live
Celebrate the holiday with exclusive offers! A new discounted group is live for high-quality generation.
New Promo Group: image 2
Enjoy 40% OFF in this special session, significantly lowering image generation costs!
✨ Includes 3 Core Models:
Nano Banana (Banana Model)
Imagen 2 (Image 2)
Imagen 2.5 (Image 2.5)
2026-09-28 15:02 (UTC+8) · #19
Section titled “2026-09-28 15:02 (UTC+8) · #19”📢 【Service Restoration Notice】 Gemini image and video generation models are now fully restored to normal service.
⚡ The platform has completed capacity expansion and channel repairs, significantly boosting generation speed and service stability.
🎉 Thank you for your patience! Feel free to submit API requests at any time to experience the latest multimodal image and video generation features!
2026-09-28 14:55 (UTC+8) · #18
Section titled “2026-09-28 14:55 (UTC+8) · #18”🍌
BANANA GROUP
Limited-Time Offer
Banana Group Site-Wide Mega Sale
40%
OFF
Ultra-Low Barrier
¥0.06 from
Ultra-low post-discount rate
Effortless tuning & testing
Context Support
2K Params
Full long-text support
Deep prompt debugging
Flagship Image Generation Models Included
Image 2
40% OFF
Image 2.5
40% OFF
HD image generation & vision multimodal features discounted
Group rate multiplier automatically updated to 0.6, effective immediately!
Try API Now →
2026-09-23 14:11 (UTC+8) · #17
Section titled “2026-09-23 14:11 (UTC+8) · #17”Maintenance Notice
⚠️ Urgent Repairs Active
Temporary Suspension of Image & Video Models
Dear users, due to upstream service fluctuations, services have been unavailable since 10:00 AM today.
Affected Models
Includes Gemini Image, Omni, and related image/video generation models
Technical Team Working on Urgent Fix
We are investigating and expediting the fix. We will notify you immediately once resolved.
🙏 We sincerely apologize for the inconvenience and appreciate your understanding and support!
2026-09-22 22:37 (UTC+8) · #16
Section titled “2026-09-22 22:37 (UTC+8) · #16”🛠️ Model List Streamlining & Welcoming Gemini 4.0 Pro Announcement
Section titled “🛠️ Model List Streamlining & Welcoming Gemini 4.0 Pro Announcement”[Architecture Streamlining] [Preparing for Gemini 4.0 Pro]
Dear users:
To simplify model invocation logic, reduce list redundancy, and fully prepare for the arrival of Gemini 4.0 Pro, the platform has streamlined and consolidated the Gemini, Omni, and Veo model series:
Key Changes Explained:
Section titled “Key Changes Explained:”- 🎯 Consolidation into Core Models:
gemini-flashgemini-imageOmni 1.1Veo 3.1
- 🎛️ Parameter-Controlled Variants: Various previous derived variant capabilities are now unified into the primary models. You can precisely control output behavior simply by passing corresponding parameters in the request JSON (please refer to the official documentation URL for parameter mappings).
- ⚡ Smooth Transition for High-Frequency Variants: To avoid impacting your daily operations, aliases for frequently used core variants will be retained for now, ensuring you don’t need to make frequent code updates.
Details: https://docs.4allapi.com/capabilities/video-parameters/
Section titled “Details: https://docs.4allapi.com/capabilities/video-parameters/”👉 Streamlined architecture, lightning-fast response. Thank you for your understanding and support!
2026-09-22 17:07 (UTC+8) · #15
Section titled “2026-09-22 17:07 (UTC+8) · #15”🔥 Grok 4.7 Flagship Model Now Live
⚡ Key Highlights & Upgrades
🧠 Leap in Logical Reasoning —— Significantly enhanced capabilities in complex coding, mathematical reasoning, and deep analysis
⚡ Ultra-Fast High Concurrency —— Direct upstream account pools delivering ultra-low latency, high throughput, and stable supply
📉 Special Discounts Supported —— Special discount groups opened simultaneously; more quota without price increases, maintaining exceptional cost-performance
👉 Full API access is now available. Feel free to explore, purchase, and test!
2026-09-16 12:20 (UTC+8) · #14
Section titled “2026-09-16 12:20 (UTC+8) · #14”🚀 Powerhouse Duo MiniMax H3 & Wan 3.0 Now Live!
Section titled “🚀 Powerhouse Duo MiniMax H3 & Wan 3.0 Now Live!”[Powerhouse Duo] [720p / 1080p Supported]
Dear Users,
The powerhouse generative AI models MiniMax H3 and Wan 3.0 are now officially integrated into our platform!
Key Highlights:
Section titled “Key Highlights:”- Dual Resolution Support: Native support for 720p HD and 1080p Full HD resolutions to suit diverse creative scenarios.
- Enhanced Performance & Visuals: Stronger prompt comprehension paired with smoother, more refined dynamic motion.
👉 API access is now available. Feel free to try them out!
2026-09-12 01:36 (UTC+8) · #13
Section titled “2026-09-12 01:36 (UTC+8) · #13”Platform Service Adjustments & New Image Models Launch Announcement
Section titled “Platform Service Adjustments & New Image Models Launch Announcement”[Policy Adjustment] [New Release]
Dear users,
Thank you for your continued support and trust in our platform. Today we have two important service updates to share with you:
1. Notice on Adjustments to Certain Low-Cost Resources
Section titled “1. Notice on Adjustments to Certain Low-Cost Resources”Due to policy changes by upstream providers, certain previously available low-cost resources have been discontinued. We sincerely regret this change; however, to best protect your interests:
- Special Discount Group Retained: The platform will continue to keep the special discount group and provide ongoing discounted rates.
- Keys Functioning Normally: Your existing API Keys can still be called normally without interruption.
2. The Reverse GPT-Image 2.5 Trio Is Officially Live!
Section titled “2. The Reverse GPT-Image 2.5 Trio Is Officially Live!”We have rolled out new low-cost reverse image models, delivering much higher detail and image quality:
- 🎨 Model Trio:
gpt-image-2.5|gpt-image-2.5-flare|gpt-image-2.5-sunburst - 🚀 Breakthrough Quality: Supports 1K image generation with lossless upscaling up to 4K.
- 💰 More Value, Same Price: Significantly improved generation quality while pricing remains unchanged!
👉 Now fully open for calls. Feel free to try out the brand new GPT-Image 2.5 series models!
2026-09-09 16:15 (UTC+8) · #12
Section titled “2026-09-09 16:15 (UTC+8) · #12”GPT-6 & gpt-image-2.5-flare Major Release Announcement
Section titled “GPT-6 & gpt-image-2.5-flare Major Release Announcement”[Flagship Model Launch] [Ultra-HD Transparent Images]
Dear users,
The platform has officially integrated the next-generation GPT-6 large language model and concurrently launched the image generation model gpt-image-2.5-flare, delivering lightning-fast response times, higher-precision 4K image rendering, and native transparent background output capabilities!
gpt-image-2.5-flare Key Parameters:
Section titled “gpt-image-2.5-flare Key Parameters:”- Supported Quality Levels (
quality):low|med|high|xhigh|max - Transparent Background Output (
background): Set to"transparent"to directly generate images with cutout transparent backgrounds
🛠️ Example Request JSON:
Section titled “🛠️ Example Request JSON:”{ "model": "gpt-image-2.5-flare", "prompt": "A puppy with a transparent background", "size": "3840x2160", "output_format": "png", "background": "transparent", "quality": "max", "response_format": "url"}2026-09-08 19:33 (UTC+8) · #11
Section titled “2026-09-08 19:33 (UTC+8) · #11”GPT-5.5 Back Online & Major Core Model Upgrades
Section titled “GPT-5.5 Back Online & Major Core Model Upgrades”[Service Restored] [Model Update]
Dear Users,
Thank you for your patience! We have completed upstream channel capacity expansion and node optimization. GPT-5.5 is now fully back online, alongside new and enhanced high-performance models:
Update Highlights:
Section titled “Update Highlights:”- GPT-5.5 Back Online: Upstream line maintenance and concurrency node expansion have been completed, restoring stable responses.
- Added Gemini 3.8 Flash: A next-generation ultra-fast inference model with significantly improved response times and logical processing.
- Omni V2V Capabilities Enhanced: Video editing now supports control via 1 original video + 1~2 reference images, making generation more accurate.
👉 Now fully open for API calls. Experience our latest model capabilities today!
2026-09-07 22:28 (UTC+8) · #10
Section titled “2026-09-07 22:28 (UTC+8) · #10”System maintenance is scheduled from 10:00 to 03:00. Please plan your schedule accordingly.
2026-09-07 14:05 (UTC+8) · #9
Section titled “2026-09-07 14:05 (UTC+8) · #9”The GPT-5.5 model (including related derivative nodes) is currently undergoing upstream line maintenance and service node capacity expansion. To ensure overall calling stability, this model is temporarily taken offline effective immediately.
🔍 Maintenance & Alternative Notice:
🛠️ Offline Notice: During maintenance, all API requests sent to gpt-5.5 will pause responding or return timeout errors.
🔄 Recommended Alternative Models: We recommend temporarily switching tasks to Gemini 3.7 Flash, Kimi K3, or GPT-4o to keep your business uninterrupted.
⏱️ Recovery Time: The operations team is fully coordinating with upstream scheduling. The service is expected to be restored as soon as maintenance is completed, followed by a separate announcement.
We sincerely apologize for any inconvenience caused and appreciate your understanding and support!
2026-09-04 17:37 (UTC+8) · #8
Section titled “2026-09-04 17:37 (UTC+8) · #8”Grok Image & Video Service Resumption Announcement
Section titled “Grok Image & Video Service Resumption Announcement”[Status: Service Restored] [Channel: Stable & High Concurrency]
Dear Users,
Grok Image and Video Generation Models are now back online!
Following a comprehensive upgrade to upstream links and distribution channels by our technical team, current channels now fully support high-concurrency continuous requests, delivering ultra-fast responses and rock-solid stability.
Key Highlights:
Section titled “Key Highlights:”- Stable & High Concurrency: Channel nodes have been significantly expanded to effortlessly handle massive traffic during peak hours.
- Full Image & Video Support: Perfectly tailored for both high-quality image generation and video generation request demands.
- Automatic Load Balancing: Sub-second seamless failover upon detecting upstream anomalies, ensuring high API success rates.
👉 You are welcome to select Grok models from the model list to experience and call them!
2026-08-31 08:45 (UTC+8) · #7
Section titled “2026-08-31 08:45 (UTC+8) · #7”🚀 Omni 1.1 Ultra-Long 40s Video & V2V Video Editing into the Multiverse Are Now Live!
Section titled “🚀 Omni 1.1 Ultra-Long 40s Video & V2V Video Editing into the Multiverse Are Now Live!”✨ The next-generation AI models are now fully integrated into the platform. All users are welcome to test them out!
🔥 Key Release Highlights
Section titled “🔥 Key Release Highlights”- 🎬 Omni 1.1 40s Video: A major breakthrough in creative duration! Video generation can now be extended up to 40 seconds.
- ⚡ Omni Video-to-Video (V2V) Editing Model Officially Released! The treasure hunt begins…
- 🎁 Limited-Time Trial Perks: The platform now supports generating a 40s video in a single call as well as video editing capabilities—placing you right into the multiverse. Come and experience ultra-long, smooth visuals!
👉 Select Omni 1.1 40s or V2V video editing from the model list right away and start your brand-new creative journey!
2026-08-26 14:32 (UTC+8) · #6
Section titled “2026-08-26 14:32 (UTC+8) · #6”Important Announcement:
Reference for Omni v2v video upload parameters:
│ Source │ OpenAI/Sora /v1/video/generations │ Gemini …:predictLongRunning │
│ Public URL │ input_reference: “https://…” │ instances[].video: “https://…” │
│ Local File │ -F input_reference=@file.mp4 ✅ │ -F input_reference=@file.mp4 ✅ Added │
│ base64 Short Video │ input_reference: “data:video/…” │ instances[].video: “data:video/…” │
2026-08-26 07:53 (UTC+8) · #5
Section titled “2026-08-26 07:53 (UTC+8) · #5”🎉 Ultra Resource Special Discounts Live & Omni V2V Officially Released!
Section titled “🎉 Ultra Resource Special Discounts Live & Omni V2V Officially Released!”🎬 Omni Video-to-Video (V2V) Editing Model Now Open!
Section titled “🎬 Omni Video-to-Video (V2V) Editing Model Now Open!”Supports advanced features such as background replacement, character swapping, and outfit changing in videos. Key supported models: 香蕉2 / Pro (Gemini Image 3.1 / Pro), Veo Ultra, Omni series. 香蕉 and Veo support 4K output.
Section titled “Supports advanced features such as background replacement, character swapping, and outfit changing in videos. Key supported models: 香蕉2 / Pro (Gemini Image 3.1 / Pro), Veo Ultra, Omni series. 香蕉 and Veo support 4K output.”🔥 Ultra Resource Exclusive Discount Rates
Section titled “🔥 Ultra Resource Exclusive Discount Rates”- 👑 SVIP Users: Enjoy 45% OFF on svip ultra resources
- 💎 VIP Users: Enjoy 40% OFF on vip ultra resources
- 👤 Default Users: Enjoy 20% OFF on default ultra resources
Limited resources available on a first-come, first-served basis!
Section titled “Limited resources available on a first-come, first-served basis!”👉 Create a new key now and select the best discounted group
1️⃣ Click the ”+ Create API Key” button at the top of the page:
2️⃣ In the configuration dialog, locate the Group option and select the corresponding discounted group (e.g., svip ultra资源 5.5折 or 默认 ultra资源 8折):
2026-08-25 16:51 (UTC+8) · #4
Section titled “2026-08-25 16:51 (UTC+8) · #4”🎬 Omni Flash v2v Video Repainting Model is Officially Live!
Section titled “🎬 Omni Flash v2v Video Repainting Model is Officially Live!”✨ The all-new Video-to-Video editing feature is now fully available!
No complex post-production required. Simply upload your video and enter simple prompts to effortlessly achieve comprehensive video repainting and transformation!
🔥 Key Highlights
Section titled “🔥 Key Highlights”- 🌄 One-Click Background Swap: Instantly switch video scenes and natural environments, from city streets to sci-fi planets
- 👤 Subject/Character Replacement: Perfectly preserve the original motion and posture while freely changing the protagonist
- 👗 Smart Outfit Change: Stylized virtual try-on and outfit redesign for effortless fashion transformation
- ⏱️ Extended Output Duration: Break through traditional generation limits with support for up to 10 seconds of continuous high-definition rendering
👉 Select the Omni Flash v2v model and unlock your new AI video creation experience today!
2026-08-22 18:56 (UTC+8) · #3
Section titled “2026-08-22 18:56 (UTC+8) · #3”⚠️ Notice Regarding Model Tuning and Recommended Alternatives
Dear Users, According to upstream notices, some models are currently experiencing brief instability due to network issues. To ensure your experience, we recommend referring to the following alternatives:
🔴 Fable 5 / Sonnet 5: (Unstable), switching recommended
🔄 Recommended Alternative 1:
🟢 Sonnet 4.6: (More stable response)
🟢 Opus 5: (Excellent logical performance)
🔄 Recommended Alternative 2:
🟢 GPT 5.6 Sol: (Efficient, stable, and outstanding)
Our technical team is actively working on the fix and will notify you as soon as stability is restored. We sincerely apologize for any inconvenience caused!
2026-08-19 21:32 (UTC+8) · #2
Section titled “2026-08-19 21:32 (UTC+8) · #2”🚀 [VIP Exclusive Appreciation Season] is officially live!
To thank our valued users, the platform is launching a limited-time 20% OFF special discount exclusive to VIP/SVIP members!
💎 VIP / SVIP Exclusive Benefits
- 🎁 Exclusive Privilege: VIP & SVIP users enjoy 20% OFF billing when calling designated channels.
- ⚙️ Configuration: Go to “Token Management” → “Add/Edit Token”, and check to add the
vip用户优享 8折group. - ⏱️ Event Duration: Subject to official price adjustments and may end at any time without prior notice. Please monitor your token validity.
👑 Standard User Upgrade Perks
- ✨ Upgrade to VIP during the promotion to instantly unlock the 20% OFF exclusive group!
🎁 The platform will continuously roll out exclusive perks from time to time for VIP and higher-tier loyal users. Thank you for your support!
We remain dedicated to delivering stable, efficient, and cost-effective AI computing power.
4ALL API
2026-08-19 16:33 (UTC+8) · #1
Section titled “2026-08-19 16:33 (UTC+8) · #1”🚀 Gemini 3.7 Model Officially Launched!
Section titled “🚀 Gemini 3.7 Model Officially Launched!”The next-generation Gemini 3.7 is now fully integrated into the platform!
A leap forward in reasoning performance and response speed, delivering a more precise and intelligent experience. ✨ Key Highlights:
- ⚡ Performance Leap: Vastly improved thinking and reasoning efficiency with faster responses
- 🧠 Enhanced Logic: Breakthrough capabilities in handling complex queries and long contexts
- 💻 All-around Assistant: Superior performance in code generation and domain-specific knowledge 👉 Select the Gemini 3.7 model now and start your efficient, intelligent journey!
Documentation history
Section titled “Documentation history”2026-08
Section titled “2026-08”- 🎉 4All API Documentation Site Launched: Quick start, API manual, capability guides, and tool integration tutorials;
- Official claude-max channel launched (full-power pure official, real model multiplier);
- The new site 4allapi.com and console api.4allapi.com are officially live, with a brand-new blue-orange visual design;
- Asynchronous video generation endpoint(
/v1/video/generations) opened: supports Veo 3.1, Sora 2, Kling v3, Omni Flash, and other models; - Image generation supports GPT-Image series 2K/4K and portrait/landscape variants.
Subsequent model launches and price adjustments will continue to be updated on this page, and important changes will also be announced in the console bulletin board.