Distilled version of SD 3.5 Large that generates high-quality images in just 4 steps, offering faster inference and reduced costs
Details
stable-diffusion-3.5-large-turbo
Ready to integrate?
Access stable-diffusion-3.5-large-turbo via our unified API.
Providers & Pricing (1)
Stable Diffusion 3.5 Large Turbo is available exclusively through Replicate, starting at $0.04/image.
replicate/stable-diffusion-3.5-large-turbo
Output
Pricing Notes (2)
- • Distilled version that generates high-quality images in just 4 steps
- • Offers faster inference and reduced costs
Stable Diffusion 3.5 Large Turbo API OpenAI-compatible
Integrate Stable Diffusion 3.5 Large Turbo into your workflow via Lumenfall’s OpenAI-compatible API to generate high-quality text-to-image outputs with minimal inference time.
https://api.lumenfall.ai/openai/v1
stable-diffusion-3.5-large-turbo
curl -X POST \ https://api.lumenfall.ai/openai/v1/images/generations \ -H "Authorization: Bearer $LUMENFALL_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "stable-diffusion-3.5-large-turbo", "prompt": "A serene mountain landscape at sunset", "size": "1024x1024" }'# Response:# { "created": 1234567890, "data": [{ "url": "https://...", "revised_prompt": "..." }] }import OpenAI from 'openai';const client = new OpenAI({ apiKey: 'YOUR_API_KEY', baseURL: 'https://api.lumenfall.ai/openai/v1'});const response = await client.images.generate({ model: 'stable-diffusion-3.5-large-turbo', prompt: 'A serene mountain landscape at sunset', size: '1024x1024'});// { created: 1234567890, data: [{ url: "https://...", revised_prompt: "..." }] }console.log(response.data[0].url);from openai import OpenAIclient = OpenAI( api_key="YOUR_API_KEY", base_url="https://api.lumenfall.ai/openai/v1")response = client.images.generate( model="stable-diffusion-3.5-large-turbo", prompt="A serene mountain landscape at sunset", size="1024x1024")# { created: 1234567890, data: [{ url: "https://...", revised_prompt: "..." }] }print(response.data[0].url)Gallery
View all 2 imagesStable Diffusion 3.5 Large Turbo FAQ
Stable Diffusion 3.5 Large Turbo starts at $0.04 per image through Lumenfall. Pricing varies by provider. Lumenfall does not add any markup to provider pricing.
You can use Stable Diffusion 3.5 Large Turbo through Lumenfall's OpenAI-compatible API. Send requests to the unified endpoint with model ID "stable-diffusion-3.5-large-turbo". Code examples are available in Python, JavaScript, and cURL.
Stable Diffusion 3.5 Large Turbo is available through Replicate on Lumenfall. Lumenfall automatically routes requests to the best available provider.
Stable Diffusion 3.5 Large Turbo supports images up to 1024x1024 resolution.
Overview
Stable Diffusion 3.5 Large Turbo is a distilled version of Stability AI’s most advanced text-to-image model. It is designed specifically for high-speed generation, utilizing adversarial diffusion distillation to produce detailed images in significantly fewer steps than the standard Large variant. This model balances the sophisticated prompt adherence of the SD3.5 architecture with the efficiency required for near-real-time applications.
Strengths
- Inference Speed: Capable of generating high-quality images in just 4 sampling steps, compared to the 30–50 steps typically required by non-distilled models.
- Text Rendering: Inherits the Multimodal Diffusion Transformer (MMDiT) architecture’s ability to render legible, coherent text and typography within images.
- Prompt Adherence: Shows high sensitivity to complex descriptors, maintaining compositional integrity even when given long or nuanced natural language prompts.
- Operational Efficiency: Lower computational overhead enables reduced costs per generation and faster iteration cycles for developers and designers.
Limitations
- Anatomical Consistency: While improved over previous versions, the model can still struggle with complex human anatomy, particularly limb positioning and finger counts in crowded scenes.
- Dynamic Range: The distillation process can sometimes lead to a slight reduction in fine-grained texture detail or a narrow dynamic range compared to the full 8-billion-parameter non-turbo model.
- Style Variation: The “Turbo” nature of the model may prioritize speed over extreme artistic flexibility, occasionally leaning toward a specific stylistic aesthetic if prompts are too brief.
Technical Background
The model is built on the Stable Diffusion 3.5 Large framework, which utilizes a Multimodal Diffusion Transformer (MMDiT) architecture featuring separate sets of weights for image and text representations. To achieve its speed, Stable Diffusion 3.5 Large Turbo employs a distillation technique that compresses the sampling process, allowing the model to bypass the standard iterative refinement of traditional diffusion while retaining the majority of its visual fidelity.
Best For
This model is ideal for interactive applications where latency is a critical factor, such as live whiteboarding tools, instant asset generation for games, and rapid prototyping workflows. It is particularly effective for users who need to iterate on visual concepts quickly without sacrificing prompt accuracy. Stable Diffusion 3.5 Large Turbo is available for exploration and integration through Lumenfall’s unified API and playground.
Try Stable Diffusion 3.5 Large Turbo in Playground
Generate images with custom prompts — no API key needed.