BizyAir IndexTTS 2.5 Text to Speech is an industrial-level zero-shot text-to-speech system that clones a voice from a single reference audio clip. It supports 5 languages (Chinese, English, Japanese, Spanish, Arabic) with cross-lingual and timbre-emotion disentanglement, fine-grained emotion control via emo_vector / emo_audio_prompt / emo_text, speaking speed control via duration_factor (0.5x–2.0x), and pronunciation control (Pinyin / CMU phonemes / Japanese Kana). Features faster inference than IndexTTS 2, BF16 inference, and vLLM production deployment. 0.8B params, returns wav/mp3/flac.

Yinchaoyongxian Song Generation is a fully self-developed Chinese AI music model that turns a one-line prompt or custom lyrics into complete songs. It supports the v4.5 model with prompts up to 3000 characters, generates up to 2 songs per call, and outputs mp3 with streaming playback. Ideal for short-video soundtracks, brand jingles, and personalized music gifts.

Yinchaoyongxian BGM Generation is a fully self-developed Chinese AI music model that turns a one-line prompt into instrumental background music without vocals. It supports the v4.5 model, generates up to 2 tracks per call, and outputs mp3 with streaming playback. Ideal for video soundtracks, podcast backdrops, and ambient scenes.

Yinchaoyongxian Reference Song Generation creates new songs that follow the style of an uploaded reference audio with controllable similarity. It works on the v3.5 model and accepts a prompt or lyrics alongside the reference track, generating up to 2 songs per call. Ideal for style-matching covers, brand audio consistency, and music prototyping.

PixVerse V6 Video Extend is AIsphere's video generation model that prolongs an existing video with consistent visuals and optional fresh audio, sharing the 360P to 1080P resolutions and up-to-15-second extension length of the V6 family. Ideal for lengthening short clips, finishing unfinished shots, and building longer narratives.

PixVerse V6 Text to Video is AIsphere's video generation model that turns text prompts into vivid video clips, supporting 360P to 1080P resolutions, five creative styles, durations up to 15 seconds, and optional synchronized audio generation. Ideal for social content, marketing creatives, and rapid idea prototyping.

PixVerse V6 Image to Video is AIsphere's video generation model that animates a single image into fluid video with prompt-guided motion, offering the same 360P to 1080P resolutions, durations up to 15 seconds, and optional audio generation. Ideal for character animation, product showcases, and animating old photos.

PixVerse V6 FLF to Video is AIsphere's video generation model that generates smooth transitions between a first frame and a last frame you provide, with 360P to 1080P resolutions, durations up to 15 seconds, and creative styles. Ideal for loop clips, scene transitions, and storyboard previsualization.

Lightricks LTX 2.5 Pro Text to Video is the premium tier delivering the highest fidelity and motion quality with native synchronized audio at 720P and 1080P, fine camera control, and 6 to 10 second quality-tuned durations. Ideal for campaign visuals, film-grade previs, and hero content.

Lightricks LTX 2.5 Pro FLF to Video generates premium motion from a first frame with optional last-frame control and native audio, the quality-first path for image-driven storytelling at 720P and 1080P. Ideal for luxury product films, key art animation, and signature brand moments.

Lightricks LTX 2.5 Fast Text to Video turns prompts into videos from 720P up to 4K with native synchronized audio. It supports 24-50 fps, cinematic camera motions, and durations up to 20 seconds at the family's cost-efficient tier. Ideal for high-volume production, previz, and rapid iteration.

Lightricks LTX 2.5 Fast FLF to Video generates a video from a first frame with an optional last frame for precise start and end composition, running from 720P to 4K with native audio at the cost-efficient tier. Ideal for product motion showcases, bringing comic panels to life, and controlled shot planning.

Google Gemini 4.0 Argon Vision is Google's flagship multimodal understanding model that processes text, image, video, and audio inputs within a unified 1M context window. It delivers state-of-the-art reasoning across modalities with long-context comprehension. Ideal for document understanding, video analysis, and multimodal assistants.

Google Gemini 4.0 Argon Text Generation is Google's flagship reasoning model with a unified 1M context window and long decodes up to 1M tokens. It excels at complex software engineering, mathematical and scientific problem solving, and autonomous multi-step agents, with output billed per token including thinking. Ideal for coding agents, deep research, and long-form content generation.

FLUX.3 Klein Text to Image is Black Forest Labs' image generation model that turns text prompts into high-quality images. It supports flexible aspect ratios, 1K and 2K resolutions, up to 4 outputs per call, reproducible seeds, and prompt expansion for richer captions. Ideal for concept art, marketing assets, and production pipelines.