Google Imagen vs Iris-3B

Google Imagen

Shutdown dates were published in the Gemini API changelog and the Vertex AI release notes, with replacements named. Discontinued. Calls to imagen model names fail on both the Gemini API and Vertex AI.

Iris-3B

Code and weights are Apache-2.0, and one Python package runs text-to-image, a 4-step Turbo, depth and 4x restoration from a documented CLI. It is ten days old, with no PyPI package, version tags, changelog or CI, needs an NVIDIA GPU and about 12 GB of weights, and has no server or API of its own.

Google Imagen

1. Replace client.models.generate_images with client.models.generate_content and the gemini-nano-banana-2.1 model, the replacement Google's deprecations table names
2. Read images from the response content parts, not from generated_images
3. Don't budget from the Vertex Imagen price rows, they describe a retired product
4. If the job needs a mask-based inpaint, use another vendor's fill endpoint, since Gemini image editing is prompt-based

Iris-3B

1. Check for an NVIDIA GPU with CUDA and Python 3.11 or newer before downloading. The base weights are about 12 GB, and Qwen3-VL-4B downloads on the first run
2. Exclude depth/*, upscaler/* and turbo/* in hf download unless those models are needed. Each of the first two adds about 12 GB
3. Don't pass --steps, --cfg-scale, --negative-prompt, --order or --shift with a Turbo checkpoint. The script exits with an error
4. Pass --seed to reproduce an image, and set --outdir, since files are named after the prompt and a repeat run overwrites them
5. Load only exported directories (model.safetensors and config.yaml). A training .pth is read with torch.load(weights_only=False), so never load one from an untrusted source