Omni Flash

Tạo video qua Labs Flow (browser automation). text→video hoặc ảnh→video (Start Image). Cần account Labs Flow ấm.

Image to video

Grok Video 1.5

Tạo video qua grok.com/imagine (browser, cần SuperGrok). Text-to-video (chỉ prompt) hoặc image-to-video (kèm 1 ảnh). Aspect Ratio áp dụng cho text-to-video; có ảnh input (i2v) thì giữ tỉ lệ ảnh gốc.

Claude Opus 4.6 (Thinking) — model mạnh nhất dòng Claude với chế độ suy luận mở rộng (thinking), xuất sắc ở suy luận phức tạp, lập trình agentic và tác vụ chuyên môn. Chạy qua gói Google AI Ultra (Antigravity).

Claude Opus 4.8

Claude Opus 4.8 delivers a major upgrade in complex reasoning, agentic coding, and professional knowledge work. With improved honesty, stronger long-horizon task execution, and a 1-million-token context window on supported platforms, it is built as Anthropic’s most capable generally available model for high-autonomy workflows and demanding enterprise use cases.

Claude Sonnet 4.6 (Thinking) — model của Anthropic với chế độ suy luận mở rộng (thinking), cân bằng tốc độ và chất lượng cho lập trình và tác vụ tri thức. Chạy qua gói Google AI Ultra (Antigravity).

GPT-OSS 120B — model ngôn ngữ trọng số mở 120 tỉ tham số của OpenAI, mạnh ở suy luận và lập trình. Chạy qua gói Google AI Ultra (Antigravity).

Gemini 3.1 Pro — model mạnh nhất dòng Gemini cho suy luận phức tạp, lập trình và phân tích chuyên sâu. Đa phương thức (ảnh/video/âm thanh). Tier High: tối đa độ sâu suy luận. Chạy qua gói Google AI Ultra (Antigravity).

Gemini 3.1 Pro (Low) — Gemini 3.1 Pro ở tier nhẹ hơn, nhanh và tiết kiệm hơn nhưng vẫn giữ năng lực suy luận tốt. Đa phương thức (ảnh/video/âm thanh). Chạy qua gói Google AI Ultra (Antigravity).

Gemini 3.5 Flash (High) — tier suy luận sâu hơn của Gemini 3.5 Flash, ưu tiên chất lượng cho tác vụ phức tạp. Đa phương thức (ảnh/video/âm thanh). Chạy qua gói Google AI Ultra (Antigravity).

Gemini 3.5 Flash (Low) — tier nhanh và tiết kiệm nhất của Gemini 3.5 Flash, hợp với tác vụ đơn giản cần phản hồi nhanh. Đa phương thức (ảnh/video/âm thanh). Chạy qua gói Google AI Ultra (Antigravity).

Gemini 3.5 Flash — model đa phương thức của Google, cân bằng tốc độ và chất lượng. Hỗ trợ nhập ảnh, video, âm thanh. Tier Medium: cân bằng độ sâu suy luận và độ trễ. Chạy qua gói Google AI Ultra (Antigravity).

Claude Opus 4.8 delivers a major upgrade in complex reasoning, agentic coding, and professional knowledge work. With improved honesty, stronger long-horizon task execution, and a 1-million-token context window on supported platforms, it is built as Anthropic’s most capable generally available model for high-autonomy workflows and demanding enterprise use cases.

Claude Opus 4.8 delivers a major upgrade in complex reasoning, agentic coding, and professional knowledge work. With improved honesty, stronger long-horizon task execution, and a 1-million-token context window on supported platforms, it is built as Anthropic’s most capable generally available model for high-autonomy workflows and demanding enterprise use cases.

Claude Opus 4.7

Claude Opus 4.7 delivers a quantum leap in logical reasoning and creative synthesis. With its next-gen architecture and massive 2-million-token context, it serves as the ultimate high-performance partner for those pushing the boundaries of innovation and discovery.

Claude Opus 4.6

Claude Sonnet 4.6 delivers strong reasoning, advanced coding capabilities, and native computer-use functionality. With adaptive thinking and a context window of up to 1 million tokens, it supports building reliable AI agents and large-scale automation workflows.

Claude Opus 4.5

Claude Opus 4.5 is Anthropic’s most intelligent model, combining maximum capability with practical performance across reasoning, coding, and complex problem solving.

GPT-5-Codex

OpenAI Codex API delivers advanced coding capabilities built on GPT-5 Codex models. It supports code generation, debugging, and complex engineering workflows with scalable reasoning depth.

GPT-5.5

GPT-5.5 is OpenAI’s advanced reasoning model for agentic coding, knowledge work, scientific research, and complex multi-step task execution.

GPT-5.4

OpenAI's GPT-5.4 delivers powerful capabilities in complex reasoning, coding, and professional knowledge work. It enables faster development cycles, more reliable task execution, and higher quality results in applications.

Gemini-2.5-Pro

Gemini 2.5 Pro is Google’s advanced thinking model designed for complex reasoning, code generation and long-context understanding. It supports native multimodal inputs and a context window of up to 1 million tokens for demanding analysis and development workflows.

Gemini-3-Flash

Gemini 3 Flash is a high-performance large language model developed by Google DeepMind, designed to deliver frontier-level reasoning with low latency and high throughput. It demonstrates strong performance across reasoning, multimodal understanding, and agentic benchmarks while supporting up to a 1M-token context window, making it well suited for real-time, high-concurrency, and production-scale AI workloads.

Gemini-3.1-Flash

Gemini 3.1 Flash API is the latest general-purpose LLM developed by Google DeepMind, designed to bridge the gap between high-speed execution and deep logic. It empowers developers to build sophisticated agents with state-of-the-art accuracy in coding, creative writing, and cross-modal analysis.

Gemini-3.1-Pro

Gemini 3.1 Pro API is the latest general-purpose LLM developed by Google DeepMind, designed to bridge the gap between high-speed execution and deep logic. It empowers developers to build sophisticated agents with state-of-the-art accuracy in coding, creative writing, and cross-modal analysis.

Gemini-3.5-Flash

Gemini 3.5 Flash is Google’s high-efficiency multimodal model, combining Flash-level speed with stronger coding, reasoning, and agentic execution for complex AI workflows that need both capability and efficiency.

Claude Sonnet 4.6

Claude Sonnet 4.6 delivers strong reasoning, advanced coding capabilities, and native computer-use functionality. With adaptive thinking and a context window of up to 1 million tokens, it supports building reliable AI agents and large-scale automation workflows.

Claude Sonnet 4.5

Developed by Anthropic, Claude Sonnet 4.5 is a premier large language model that excels in multi-step reasoning, complex coding, computer use, and specialized domains such as finance.

Claude Haiku-4-5

Claude Haiku 4.5 is Anthropic’s lightweight AI model designed for fast, cost-efficient AI applications. It delivers near-frontier performance for coding, chat, and automation tasks while providing significantly faster responses and lower cost than larger Claude models.

Wan 2.6 Image (Edit)

Multi-image editing with wan2.6-image by Alibaba (DashScope): 1-4 reference images -> 1-4 edited images. Style transfer, subject consistency. Direct connection, no gateway.

Wan 2.7 Interactive Edit

Interactive editing with wan2.7-image-pro by Alibaba: place/blend objects using bounding boxes (bbox_list). Up to 2K. Direct connection, no gateway.

Wan 2.7 Image Edit

Image editing with wan2.7-image-pro by Alibaba (Wan 2.7 / DashScope). Attach 1-3 images + an instruction. Up to 2K. Direct connection, no gateway.

Qwen Image Edit

Sửa ảnh theo chỉ dẫn ngôn ngữ bằng Qwen-Image-Edit của Alibaba (DashScope). 1 ảnh (instruction-only) hoặc nhiều ảnh (reference). Kết nối thẳng, không qua gateway.

Image to video

Nano Banana Pro

Google DeepMind’s Nano Banana Pro delivers sharper 2K imagery, intelligent 4K scaling, improved text rendering, and enhanced character consistency—offering a major leap in visual quality for creative and API-driven workflows.

Nano Banana 2

Google DeepMind’s Nano Banana 2 delivers sharper 2K imagery, intelligent 4K scaling, improved text rendering, and enhanced character consistency—offering a major leap in visual quality for creative and API-driven workflows.

GPT Image 2(img2img)

GPT Image 2 is OpenAI’s next-gen image model built for stronger photorealism, cleaner image editing, sharper text rendering, and more polished product photography. Designed for more advanced visual workflows, it pushes image generation beyond basic text-to-image output and into higher-quality creative, commercial, and design-ready use cases.

GPT Image 2

GPT Image 2 is OpenAI’s next-gen image model built for stronger photorealism, cleaner image editing, sharper text rendering, and more polished product photography. Designed for more advanced visual workflows, it pushes image generation beyond basic text-to-image output and into higher-quality creative, commercial, and design-ready use cases.

Grok Image

Tạo/sửa ảnh qua grok.com/imagine. Text-to-image (chỉ prompt → 4 ảnh) hoặc image-to-image (kèm 1–3 ảnh input → 1 ảnh sửa). Aspect Ratio áp dụng cho text-to-image; khi có ảnh input, output GIỮ tỉ lệ ảnh gốc (aspect bị bỏ qua). Cần gói SuperGrok.

Imagen 4

Tạo ảnh bằng Imagen 4 của Google qua Labs Flow. text→ảnh hoặc ảnh→ảnh (1–4 ảnh tham chiếu). Cần account Labs Flow ấm.

Seedream 4.5

Seedream 4.5 delivers better editing consistency, improved multi-image fusion, finer detail control, natural small text and faces, and harmonious, aesthetic visuals.

Seedream 5.0

The latest image generation model released by BytePlus. Further upgraded reasoning and comprehension capabilities, accurately interpreting complex prompts and visual inputs.

Seedream 4.0

Seedream 4.0 is a SOTA multimodal image generation model, enabling text-to-image creation, image editing and multi-image generation within a single model.

Wan 2.7 Image Set

Coherent image-set generation with wan2.7-image-pro by Alibaba: one prompt -> N consistent images (enable_sequential). Up to 2K. Direct connection, no gateway.

Wan 2.7 Image

Text-to-image with wan2.7-image by Alibaba (Wan 2.7 / DashScope). 2K resolution. Direct connection, no gateway.

Wan 2.7 Image Pro

Text-to-image with wan2.7-image-pro by Alibaba (Wan 2.7 / DashScope). 4K resolution. Direct connection, no gateway.

Qwen Image Plus

Text-to-image with qwen-image-plus by Alibaba (Qwen/DashScope). Optional reference images. Direct connection, no gateway.

Qwen Image Max

Text-to-image with qwen-image-max by Alibaba (Qwen/DashScope). Optional reference images. Direct connection, no gateway.

Qwen Image 2.0 Pro

Text-to-image with qwen-image-2.0-pro by Alibaba (Qwen/DashScope). Optional reference images. Direct connection, no gateway.

Qwen Image 2.0

Text-to-image with qwen-image-2.0 by Alibaba (Qwen/DashScope). Optional reference images. Direct connection, no gateway.

Wan 2.6 Image (Interleaved)

Text-to-image with interleaved text+image output via wan2.6-image by Alibaba (DashScope, SSE streaming reassembled server-side). Optional 1 reference image. Direct connection, no gateway.

text-to-speech

Omni Flash

Tạo video qua Labs Flow (browser automation). text→video hoặc ảnh→video (Start Image). Cần account Labs Flow ấm.

Grok Video 1.5

Tạo video qua grok.com/imagine (browser, cần SuperGrok). Text-to-video (chỉ prompt) hoặc image-to-video (kèm 1 ảnh). Aspect Ratio áp dụng cho text-to-video; có ảnh input (i2v) thì giữ tỉ lệ ảnh gốc.