Comet API Blog

The CometAPI Blog shares practical guides and updates on mainstream

AI models to help developers get started quickly and integrate them efficiently.

How to Use Midjourney’s V1 Video Model?

Midjourney shook the AI art community in mid-June 2025 by unveiling its inaugural Video Model, V1, marking a significant expansion from static image generation into animated content. This long-anticipated feature was officially announced on June 18, 2025, via Midjourney’s blog, with broad accessibility granted on June 19, 2025 . In practical terms, V1 allows creators […]

How to Download Suno AI Songs

In an era where AI-generated music is rapidly reshaping the industry, Suno AI stands out as one of the most popular platforms for creating original songs from simple prompts. If you’ve crafted a song you’re proud of, the next step is downloading it—whether for personal listening, sharing on social media, or integrating into your own […]

Alibaba Cloud Unveils Qwen‑TTS: A High‑Fidelity, Streaming Speech Synthesis Model

On June 26, 2025, Alibaba Cloud launched Qwen‑TTS, the latest addition to its Tongyi Qianwen (Qwen) family of large AI models. Designed for versatile, high‑quality text‑to‑speech applications, Qwen‑TTS supports Chinese, English, and mixed‑language input and offers both batch and streaming audio outputs, catering to diverse use cases from intelligent voice assistants to multimedia content production. […]

Is o3‑mini Out? An In-depth Analysis

In early 2025, OpenAI introduced o3‑mini, a compact yet powerful “reasoning” model designed to deliver high-performance results in STEM tasks at reduced cost and latency. Since its public debut on January 31, 2025, o3‑mini has been integrated into ChatGPT’s model picker and made accessible via API to developers and end users under various plan tiers. […]

Is Sora Available to the Public

Sora, OpenAI’s groundbreaking text-to-video model, has captured global attention since its debut in early 2024. Originally introduced as a research preview, the technology promised unprecedented capabilities—generating high-definition videos purely from text prompts. Over time, it moved from closed testing to broader access, sparking questions about who can use it, where it’s available, and what its […]

Alibaba Cloud releases Qwen‑VLo multimodal model,Image capability upgrade

Alibaba Cloud’s AI division has officially launched Qwen‑VLo, the latest iteration in its Qwen multimodal model series, marking a significant advancement in unified vision‑and‑language capabilities. Announced on June 28, 2025, Qwen‑VLo offers both understanding and generation functionalities, extending well beyond its predecessors to include high‑resolution image creation and editing driven by natural‑language prompts and visual […]

Can Claude Create Images? All You Need to Know

In recent months, a growing number of developers and enterprises have asked a common question: Can Anthropic’s Claude models generate new images directly? While Claude has made impressive strides in multimodal understanding—allowing users to upload and analyze images—the ability to natively generate novel visuals remains a point of confusion. What is Claude and what can […]

ChatGPT Shopping — Here’s What is and How to Use!

With the rapid evolution of large language models (LLMs), ChatGPT has moved from a purely conversational AI to a fully capable shopping assistant. Over the past two months, OpenAI has rolled out native in‑chat shopping features—powered by partnerships with Shopify, Visa, and major retailers—allowing users not only to explore and compare products, but to complete […]

What is AI Hallucination?

What Is AI Hallucination? AI hallucination refers to the phenomenon where artificial intelligence models—especially large language models (LLMs) and generative AI systems—produce outputs that are plausible in form but contain false, fabricated, or misleading information. These “hallucinations” can range from the invention of fictitious facts and citations to erroneous interpretations of user queries. While such […]

What is Deepthink R1? All You Need to Know

DeepSeek, a fast-rising Chinese AI firm, recently launched DeepThink R1, an advanced reasoning model built atop its popular R1 series. The model has quickly made headlines—earning comparisons to OpenAI’s top models, dominating benchmarks, and attracting global attention. This article delves into DeepThink R1: what makes it special, how it fits into DeepSeek’s R1 lineage, its […]

1 8 9 10 11 12 45