​
Model List

Most supported models are open source. View public models here

You can use the following model name as model_id in Heurist API/SDK.

​
Large Language Models (LLMs)

  • mistralai/mixtral-8x22b-instruct: state-of-the-art Sparse Mixture-of-Experts model featuring 8 experts and 22 billion parameters, optimized for instruction-following with strong performance in reasoning and multilingual capabilities.

  • mistralai/mixtral-8x7b-instruct: a compact version of the Mixtral family, designed for efficient instruction-based interactions while maintaining a balance between performance and cost.

  • mistralai/mistral-7b-instruct: excel in generating fast responses and highly cost-efficient.

  • meta-llama/llama-3.1-405b-instruct: the most advanced Llama model, perfect for high-level reasoning and complex task execution.

  • meta-llama/llama-3-70b-instruct: the most popular Llama model, suitable for diverse applications in conversational AI and reasoning.

  • nvidia/llama-3.1-nemotron-70b-instruct: specialized version of the Llama model tailored by NVIDIA for complex instruction-following tasks, delivering high-quality, human-like responses across a variety of applications while leveraging advanced NVIDIA AI technologies for optimal performance and scalability.

  • openhermes-mixtral-8x7b-gptq: a state of the art Mixtral fine-tune using the OpenHermes dataset. Uncensored

  • hermes-3-llama3.1-8b: flagship Hermes LLM trained by Nous Research, with advanced agentic capabilities, enhanced roleplaying, reasoning, multi-turn conversation, long context coherence and agentic abilities. Uncensored. Supports tool calling

  • dolphin-2.9-llama3-8b: an enhanced Llama3 8b model trained by Cognitive Computations equipped with a variety of instruction, conversational, and coding skills. We recommend using system prompt The assistant is named Dolphin. A helpful and friendly AI assistant, Dolphin avoids discussing the system message unless directly asked about it.. Uncensored.

  • theia-llama-3.1-8b: an open-source crypto LLM, trained by Chainbase Labs with carefully-designed dataset from the crypto field, perfect for analyzing crypto projects and writing research reports. Learn more about Theia

​
Image Generation Models

  • HeuristLogo: Flux LoRA that can generate the logo of Heurist. Trigger word: Heuristai logo or hexagonal logo.

  • FLUX.1-dev: state-of-the-art open source image generation model that excels at a variety of image styles.

  • PepeXL: SDXL LoRA that can generate the meme frog, Pepe. Trigger word: Pepe. Recommended words: pepe_frog, meme, web comic and add cartoon or 3d render

  • Zeek: SDXL LoRA for ZKsync’s mascot cat, Zeek. Trigger word: zeek. We recommend using zeek wearing ... or zeek disguised as ...

  • MagusDevon: SD1.5 LoRA for Magus Devon, the witch character of Forgotten Runes. Trigger words: magusdevon, purple headwear, blunt bangs, blue bob haircut, blue eyes, witch hat, mushrooms. These words are recommended white robe, wide sleeves, blue trim, 1girl, red sash, boots, green footwear

  • Aurora: SD1.5 checkpoint for anime girls.

  • AnimagineXL: SDXL checkpoint for anime images. It can generate characters from well-known anime series.

  • CyberRealisticXL: SDXL checkpoint for realistic portraits.

  • BrainDance: SD1.5 checkpoint for cartoon, anime and watercolor styles.

  • BlazingDrive: SD1.5 checkpoint for anime characters and scenes.

  • YamersCartoonArcadia: SD1.5 checkpoint for stylized 2D cartoon.

  • ArchitectureRealMix: SD1.5 checkpoint for architectural design, landscape design, urban design, and interior design.

  • ArthemyComics: SD1.5 checkpoint for fantasy cartoon images.

  • ArthemyReal: SD1.5 checkpoint for fantasy-style and realistic-looking images.

  • DreamShaperXL: SDXL checkpoint that aims at doing everything well, including photos, art, anime, and manga styles.

  • AAMXLAnimeMix: SDXL checkpoint for anime art and hentai.

  • SDXLUnstableDiffusersV11: SDXL checkpoint that enhances SDXL capabilities in creating vibrant arts, designs and photo-realistic images.

  • SDXL: general-purpose image generation model developed by Stability AI.

​
Understanding Image Prompting

Stable Diffusion can benefit significantly from the use of comma-separated words and phrases in its prompts. Using commas effectively organizes prompts into distinct elements, enhancing clarity and aiding in the generation of more relevant images. For example, a prompt structured as “a cat, a dog, a tree” allows the model to interpret each item separately, which can lead to more accurate and coherent outputs compared to a continuous paragraph of description. This practice is especially useful for creators who want to ensure that specific attributes or elements are emphasized in their generated images.

Flux models work best if the prompt contains complete sentences depicting the image. Comma-separated tags are not recommended. This is a major difference between Stable Diffusion and Flux.

LoRA is a fine-tuning technique that makes a model learn a specific concept, style, item, or character. LoRA requires “trigger words” to function. For example, the phrase “Heuristai logo” must be included if you want the HeuristLogo model to generate a correct logo. Some LoRA models require multiple words or phrases to trigger effectively.

We recommend visiting Heurist Imagine and see the example prompts of the models that you want to use.

​
Need More Models?

Any image generation models from Civitai and LLMs from HuggingFace can be supported upon request. If you’re interested in hosting your models on Heurist, or if you want customized models that adapts to your specific use cases, please contact us at team@heurist.xyz

Integration OverviewImage Generation API