✅ Website 502 from 23:07–23:20 (UTC+8) tonight is resolved — please retry tasks that failed during the window —— An upstream backend server vendor started its maintenance (originally scheduled for 7/23 01:00–05:00 UTC+8) roughly 2 hours early, causing a sudden interruption; our backend failover was ready to switch at any time, but cache and DNS resolution caching still caused about 15 minutes of disruption. Force-refresh if the site feels sluggish — API endpoints are no longer affected.📖 View details
🚀 Gemini 3.6 Flash & 3.5 Flash-Lite dual launch, 30+ test cases verified per model on both endpoints —— Google’s two July stable text models are live in the
default / svip groups: gemini-3.6-flash at $1.50/$7.50 and gemini-3.5-flash-lite at $0.30/$2.50 per 1M tokens, identical to official pricing; Search grounding, code execution and other native tools all verified working — 3.6 thinks by default (four controllable tiers), Lite ships zero-thinking with ~2s responses. Onboarding: 3.6 Flash Overview.📖 View details🚀 HappyHorse video series now on sale: flagship upgraded to 1.1, default 98% of official price —— Four Alibaba video models in one release:
happyhorse-1.1-t2v / happyhorse-1.1-i2v / happyhorse-1.1-r2v (up to 9 reference images) / happyhorse-1.0-video-edit; shares the Wan&HappyHorse group and endpoint with Wan — just switch the model name. Pricing is built into the system: 0.14x rate × fixed 1:7 exchange rate = 98% of official, down to ~81% with recharge bonuses. Onboarding: HappyHorse Overview.📖 View details🚀 Seed 2.1 Turbo is live — both endpoints verified 15/15 —— ByteDance’s high-frequency production text model
dola-seed-2-1-turbo-260628 is available on the default / svip groups at $0.50 input / $2.50 output per 1M tokens; Chat Completions and Responses are both verified, with implicit + explicit two-layer caching. Note deep thinking is ON by default — disable it explicitly for high-frequency workloads. Onboarding: Model Overview.📖 View details🚀 AI Image Master playground upgraded: hardened network stability and a redesigned model picker UI —— The model picker is now grouped by vendor (Google / OpenAI / Seedream / Flux — 17 models), with each model labeled with its expected generation time and per-image price; network-layer stability has been hardened in the same release. Try it login-free: imagen.apiyi.com.📖 View details
🚀 Kimi K3 is live — $3/$15 per 1M tokens, identical to official pricing —— Moonshot AI’s 2.8-trillion-parameter open-source flagship released 7/16, the largest open-source model ever: a 1M context priced flat across the whole range, #1 on LMArena Frontend Code Arena, native image/video input; APIYI has launched
kimi-k3, with cache-hit input as low as $0.30 and recharge bonuses giving an easy ~9% off (recharge $100, get 10%), up to ~17% off — deep dive at Kimi K3 Launch.📖 View details✅ Nano Banana Pro & 2 run on pure official AI Studio relay — consistently stable, bring your traffic —— Since Nano Banana launched in November 2025,
gemini-3-pro-image / gemini-3.1-flash-image and the lite series have run exclusively on the official AI Studio relay (an official Vertex group is also available) — never a reverse-engineered channel; a price well below official does not mean a reverse-engineered source. Professionalism earns trust.📖 View details✅
gpt-image-2-all speed is back to normal — fast cases complete in about 40 seconds —— The ChatGPT website has recovered its compute capacity and this reverse-engineered model’s speed has returned to normal with it, with first-byte times mostly in the 30-60 second range in our logs; send traffic with confidence — no RPM cap to worry about, and supply is ample.📖 View details📊 Nano Banana Pro & 2 have stayed fast for the past 20 days — averaging about 30 seconds per image —— Cards on the table:
gemini-3-pro-image and gemini-3.1-flash-image, the core workhorse image models on APIYI, have been consistently fast in our monitoring over the past 20 days and just as stable under high concurrency; we’ll keep working hard to safeguard operations and supply.📖 View details📚 New docs group “Image API Essentials” with the new “Must-Read & Best Practices” page now live —— Key takeaway: all image models are synchronous calls with no async task ID — a dropped connection loses the result but is still billed, so a generous
timeout is rule number one; the page includes per-model timeout recommendations and a base64 / URL output reference table, and the existing “Image Compression & Output Resolution” and “How to Generate Satisfying Images” pages have moved into this group — read it at Must-Read & Best Practices.📖 View details📖 For earlier updates, visit the Live Updates Archive — browse history by month, category, or vendor.
Deep Dive
AI Radar
Telegram
Global users
Media Models
New additions