What is Blaze AI
Blaze AI is a standalone visual creation studio within the ProMediaFlow platform designed for fast, single-step AI media creation. It enables creators to generate high-resolution images, cinematic video clips, custom illustrations, mood tests, and pitch visuals directly—without setting up or configuring formal production projects.
Blaze AI operates across three main workspace tabs: Image, Video, and Chat. It aggregates over 30 AI models across 7 provider backends (Google, Black Forest Labs, BytePlus Ark, Kling AI, MiniMax, OpenRouter, and Groq). Generations render synchronously with real-time progress updates and immediate delivery of completed files to your personal library (My Work). From your library, you can inspect, edit, touch up, download, share, or publish your creations to the community gallery.
Getting started
You can access Blaze AI and generate media from desktop or mobile devices:
- Open the Application: Select Blaze AI from the main ProMediaFlow navigation menu.
- Choose a Workspace Tab: Select Image, Video, or Chat from the workspace tabs at the top of the screen.
- Select Model and Quality Tier: Open the configuration panel to select your preferred AI model provider and quality tier (such as Free, Pro, or Ultra).
- Enter Prompts and References: Type your text prompt into the composer input field. You can also drag and drop reference images directly into the composer to guide style, character consistency, or framing.
- Configure Parameters & Burst Count: Set your aspect ratio, output resolution, and select how many variations to produce per submission (1 to 4 outputs). On mobile devices, tap the Settings chip to open the settings bottom sheet and access the count picker.
- Review Live Pricing or Run a Preview: Check the live price calculation near the send button before initiating your request. AI generations consume account credits based on your chosen model and parameters. You can click to run a zero-credit dry-run preview to test parameters and verify cost estimates without deducting credits or rendering files.
- Generate: Click the generation button. Account credits for the request are validated and held prior to dispatching; if your account balance is insufficient, a short-balance banner appears and processing halts immediately before any billing occurs.
Key capabilities
1. Text-to-Video Synthesis
Generate high-resolution cinematic video clips directly from text descriptions using a consolidated video settings panel that supports multiple aspect ratios (including 21:9 ultrawide):
- Google Veo 3.1: Renders resolutions up to 4K with clip durations of 4–8 seconds.
- Kling 3.0: Renders resolutions up to 1080p with clip durations of 3–15 seconds.
- Hailuo 2.3: Renders resolutions up to 1080p with clip durations of 6–10 seconds.
- BytePlus Seedance 1.5: Renders resolutions up to 1080p with clip durations of 4–12 seconds.
- Omni Flash 1.2: Delivers cinematic video generation with optimized request stability.
2. Image-to-Video Animation, Keyframe Controls & Audio
Animate static images into video clips with visual tracking and frame consistency:
- First-Frame Control: Upload a starting image frame across all supported video models to guide the animation sequence.
- Keyframe Interpolation & Transformations (Google Veo 3.1 Exclusively): Supply both starting and ending frames to control motion progression, attach up to 3 visual reference assets to guide video style, or perform full video-to-video transformations.
- Synchronized Audio Generation: Automatically generate and embed synchronized sound effects or audio tracks into video renders using Google Veo 3.1 or BytePlus Seedance 1.5 models.
3. Text-to-Image Generation
Produce photorealistic photos, digital artwork, and custom illustrations in a dedicated creation workspace:
- Supported Models: Choose from FLUX.2 (Flex, Pro, and Max tiers), Google Imagen 4, Nano Banana, and BytePlus Seedream (versions 4.0, 4.5, and 5.5).
- Workspace Controls: Organized layout segregating basic and advanced settings panels, negative prompt fields, face reference controls, custom LoRA parameter support, output resolutions up to 4K, and over 10 aspect ratios (including 21:9 ultrawide). Responsive prompt inputs and refined helper panels provide immediate feedback during creation.
4. Reference-Guided Image-to-Image Creation
Guide image generation by dragging and dropping reference images directly into the composer:
- FLUX.2 Models: Support up to 8 reference images per generation request.
- BytePlus Seedream & Nano Banana Pro: Support up to 14 reference images per generation request.
- Creative Guidance: Use reference images for style transfer, character consistency, or composition guidance across all image models.
5. Multimodal AI Chat & Visual Analysis
Engage in multi-turn conversations for prompt refinement, creative brainstorming, and technical evaluation in the Chat tab:
- Conversational Models: Interact with Gemini 2.5 Pro, Gemini 3 Pro, Gemini Flash series, or Gemma 3 27B with streaming text responses.
- Vision Analysis: Attach image files directly to chat prompts to analyze visuals, request technical evaluations, or ask context-aware questions.
- Conversation Management: Organize message threads in the sidebar, inspect visual assets in a specialized media lightbox viewer, and review prompt iterations using a visual version timeline.
6. In-Chat Image Generation & Iteration
Generate and iterate on images directly inside conversation threads using Nano Banana and Nano Banana Pro models:
- Switch seamlessly between text conversation and image generation modes without leaving the Chat tab.
- Reuse any AI-generated image from your chat history immediately as a reference image for follow-up prompts.
7. Multi-Generation Burst Control
Queue and generate up to four outputs simultaneously from a single prompt submission using the Generations per send picker (values 1, 2, 3, or 4; default is 1):
- Burst Creation: Produce multiple variations of the same prompt and parameter set in a single submission, creating individual library tiles for each output.
- Pre-Flight Price Breakdown: The send button displays the complete price breakdown formatted as
N × price = total before dispatching. Credits for the entire burst are validated and held prior to execution.
- Mobile Access: Access the picker from the Settings bottom sheet (via the Settings chip) on mobile devices. Your choice is remembered locally for future sessions.
8. Integrated Image Editor & Project-Less Versioning
Modify and touch up generated images directly inside Blaze AI without needing external software:
- Canvas Refinements: Adjust colors, apply style conversions, or perform touch-ups directly on the image canvas.
- Project-Less Saving: Save canvas edits, color conversions, or Topaz enhancements on your own images as new version records directly in your personal library without requiring a formal project.
9. Multi-Provider Model Hub & Live Pricing
Access over 30 AI models aggregated from 7 provider backends (Google, Black Forest Labs, BytePlus Ark, Kling AI, MiniMax, OpenRouter, and Groq):
- Live Cost Estimation: Review pre-flight cost calculations before initiating any generation.
- Model Comparison: Switch between providers and models instantly to compare visual styles and manage budgets.
- Budget Control: Access detailed pricing lists and usage summaries within the application, utilizing mobile-optimized controls to manage quality tiers on the go.
10. Generation History & Search ("My Work")
All generated visual assets save automatically to your personal library workspace:
- Interactive Masonry Grid: Browse past creations on an adjustable, date-grouped timeline with lazy-loading infinite scrolling.
- Search and Filter: Search outputs by prompt keywords, filter by AI model, or toggle between your personal library feed and the community feed.
11. Fullscreen Preview, Metadata & Gallery
Click any visual asset in your history or gallery to open an immersive fullscreen preview overlay:
- Navigation & Metadata: Navigate through generated assets using keyboard arrow keys, inspect complete prompt details, download full-resolution originals, and manage your library.
- Social Engagement: Like and share assets directly from the preview overlay.
12. Community Gallery & Sharing
- Public Sharing: Share your best generations to the public community gallery with a single click.
- Inspiration Feed: Browse a community feed of AI-generated images and videos to discover popular prompts, creative styles, and like works created by other users.
How to use it
Generating an Image
- Click the Image tab at the top of Blaze AI.
- Choose your model (e.g., FLUX.2 Pro, Google Imagen 4, or BytePlus Seedream 5.5) and set your quality tier.
- Select your aspect ratio (e.g., 21:9 ultrawide or 16:9) and resolution (up to 4K).
- (Optional) Drag reference images into the composer field for style or character consistency (up to 8 for FLUX.2, up to 14 for Seedream or Nano Banana Pro).
- Enter negative prompts or configure face reference controls in the advanced settings modal if needed.
- Set the Generations per send picker (1–4) to choose how many variations to create.
- Check the price calculation (
N × price = total) or click dry-run preview to test parameters.
- Click the generate button to create your image. Results will appear as new tiles in your personal library.
Generating a Video
- Click the Video tab.
- Select a video model (Google Veo 3.1, Kling 3.0, Hailuo 2.3, BytePlus Seedance 1.5, or Omni Flash 1.2).
- Set clip duration, aspect ratio, and resolution (up to 4K on supported models).
- (Optional) For image-to-video animation, upload a starting frame image. If using Google Veo 3.1, you can also supply an ending frame for keyframe interpolation, add up to 3 style reference assets, or toggle synchronized audio generation.
- Review the live cost estimate on the send button and click generate.
Using Multimodal Chat
- Click the Chat tab.
- Choose a conversational model (e.g., Gemini 3 Pro or Gemma 3 27B).
- Type a prompt or upload an image for visual analysis and evaluation.
- To generate an image inside chat, switch to image mode and select Nano Banana or Nano Banana Pro.
- Use any generated image in your chat history directly as a reference image for your next prompt.
- Use the sidebar to manage message threads, open the media lightbox to inspect visuals, or review prompt history using the version timeline.
Editing and Enhancing Images
- Open an image in your personal library (My Work).
- Launch the integrated image editor to adjust canvas colors, apply style conversions, or perform touch-ups.
- Click Save as new version or apply Topaz enhancement. The updated image will save as a new linked version record directly in your library without requiring a project.
Tips & notes
- Credit Validation & Pre-Flight Holds: All AI generations consume credits based on the chosen model, resolution, duration, and burst count. Account credits are validated and held before dispatching requests. If your balance is insufficient, execution halts immediately before any billing occurs.
- Zero-Credit Dry Runs: Use the dry-run preview feature to test parameter configurations and estimate exact credit costs before generating actual files.
- Burst Generations: When setting "Generations per send" to 2, 3, or 4, each output runs as part of a queue and produces an individual tile in your personal library.
- Synchronous Output Delivery: Blaze AI processes renders synchronously with live progress updates, delivering finished files straight to your library. If a video generation fails, it is explicitly marked as failed so processing loops do not persist.
- Project-Less Workflow: Blaze AI operates independently of formal production projects. Edits, color conversions, and Topaz enhancements on your own images save directly as new versions within your personal library.