Sayak Paul has published an analysis of key-value caching in flow models for image generation, detailing implementation approaches, benchmarks, and experimental findings.
Users can now run Nano Banana 2.1 workflows in ComfyUI with configurable thinking levels, up to 14 reference images, and outputs up to 4K resolution.
OpenRouter has published pricing and capability details for Nano Banana 2.1, setting generation rates from $0.0336 per image and supporting up to 14 reference images alongside search grounding.
OpenRouter has added GoogleAIDevs' Nano Banana 2.1 to its platform, bringing the updated image generation and editing model to users.
Google has announced Nano Banana 2.1, an upgraded image generation and editing model that reports improvements in visual design, mask-based editing, and subject consistency.
Codex CLI 0.158.0 is out with copy-on-select and right-click paste in the TUI, OAuth client secrets for MCP servers, and transparent backgrounds for image generation.
Runway MCP can be connected directly to Claude, including Claude Opus 5.5, to generate images and videos with Gen-4.5, Seedance 2.5, GPT Image 2, Kling, and other models.
ComfyUI says Recraft V4.1 Flash generates an image in 1.3 seconds and calls it the fastest image model on the market. A selected result can then be refined to 2048×2048 Pro quality.
The --tile update adds seamless patterns in v8.1 and v8.2, and the new inpainting changes only pixels a user selects. Real-time models are an early preview on Midjourney’s alpha site, not a full release.
GMI Cloud is offering the Hy Image 3.5 preview at $0.024 per image, a rate it says is 8.8x cheaper than GPT Image 2 and 5.6x cheaper than Nano Banana Pro.
OpenRouter lists Recraft V4.1 Flash at $0.007 per image and 1.3 seconds per image. It says the output still looks like Recraft for photography, portraits, complex scenes, logos, and early concepts.
Microsoft AI puts that pace at 377 points a year, or 1.6 times the next-fastest climb on the board.
Microsoft AI says MAI gained 283 Elo on text-to-image in 10 months, putting its image-generation hill-climbing pace 61% ahead of any other lab while optimizing the Pareto frontier.
OnSoloAI says the exclusive Hy Image3.5 preview supports 5 refs and 2K, with two weeks of free access for Members. Tencent Hy describes it for short-drama character sheets, full-motion video game assets, and keyframes.
ComfyUI has made the Hy Image3.5 preview available for text-to-image and image-to-image generation at up to 2K. ComfyUI says it has a +30% win rate against Hy Image3.0 in human evaluation.
Tencent Hunyuan said Hy Image3.5 preview is live, with text-to-image and image-to-image generation up to 2K. Tencent Cloud API pricing is $0.024 per generated image, and it offered two weeks free only on OnSolo and Miora.
Sayak Paul described how KV caching is incorporated in QwenImage 2.1 by caching fixed context separately from image positions that change at each denoising step. He said later steps become cheaper and a 2.55x speedup was measured.
Alibaba Cloud said Qwen-Image-2.1 improves visual quality over its predecessor, particularly in the aesthetics of text and portraits, and shared text-rendering examples.
Alibaba Cloud announced Qwen-Image-2.1, a unified 7B image generation and editing model with open weights. The company says it natively handles RGBA layers, supports up to 10 reference images, and is available on GitHub, ModelScope, and Hugging Face.
Qwen highlighted a Hugging Apps live demo of Qwen-Image-2.1 on Hugging Face Spaces, saying one checkpoint covers generation and editing and can be tried in the browser with no setup. Hugging Apps described the model as a 7B-parameter native generation and editing model with up to 10 image references.
