New models, improvements, and fixes. Every update, documented.
No updates yet.
Runway Gen-3 Alpha is live on Fluxpool. Text-to-video and image-to-video, up to 1080p, 10-second clips. Access it via the playground or API with model ID runway-gen3. Pricing: from $0.50/video.
Flux 1.1 Pro generation speed improved by ~30%. Average time-to-image now ~2.1s (previously ~3s) at 1024×1024. No API changes required.
Fixed an issue where webhook callbacks for video generations could be sent multiple times under high load. Webhook deliveries are now idempotent. If you implemented deduplication on your end, no changes needed.
Alibaba's latest image model is now available. Strong Asian aesthetics, excellent text rendering, multilingual prompt support. Model ID: qwen-image-3-6-plus. From $0.02/image.
Improved compatibility with the official OpenAI Python SDK v1.30+. The images.generate() and images.edit() endpoints now fully match the OpenAI response schema. Existing integrations should see no breaking changes.
Kuaishou's Kling v2 is live. Realistic motion, strong character consistency, 5-10 second clips at up to 1080p. Model ID: kling-v2. From $0.20/video.
The generations endpoint response now wraps output URLs in a data[] array to match the OpenAI format. If you were accessing response.url directly, update to response.data[0].url. See the migration guide.
New endpoint: GET /v1/credits/balance. Returns your current credit balance and usage for the current billing period. Useful for monitoring spend in automated pipelines.
Alibaba's Wan Video 2.1 is available for text-to-video and image-to-video generation. High-quality cinematic output at up to 1280×720. Model ID: wan-video-2.1. From $0.15/video.
Fixed an issue where negative_prompt was being ignored for SDXL 1.0 generations when using the OpenAI SDK format. Negative prompts now correctly apply across all image models.
Fluxpool launches with Flux 1.1 Pro by Black Forest Labs. Best-in-class image generation, up to 2048×2048. Model ID: flux-1.1-pro. From $0.03/image.
We're live. One API, multiple image and video generation models, pay-per-generation pricing. Start with 50 free generations.
Get started →