Meet Ethos 6 and Image 2: Two New AI Models from Atomesus

Today we're launching two new AI models: Ethos 6, our flagship multimodal model, and Atomesus Image 2, our in-house image generation model. Ethos 6 understands text, documents, images and video. Image 2 turns a single sentence into a finished visual. Both are live on Atomesus now, and developers can use Ethos 6 through an OpenAI-compatible API starting at $0.52 per million input tokens.

Ethos 6: our most capable model yet

Ethos 6 is built to handle the work you do every day, in one model. Ask a question, share a file or describe a task, and Ethos 6 responds in clear, natural language. It runs on our own serving infrastructure, which keeps it fast and affordable.

What Ethos 6 can do:

• Writing. Draft emails, essays, reports and posts, with context that carries across follow-up questions.
• Coding. Generate, explain and debug code across popular programming languages.
• Research. Get sourced answers and clear summaries on almost any topic.
• Reasoning. Work through multi-step problems in maths, science, finance and planning.
• Agents. Hand off multi-step tasks and let Ethos 6 carry them through end to end.

Ethos 6 can see: documents, images and video

Ethos 6 is multimodal. Upload a file, ask a question in plain words, and it reads the content for you.

• Documents (PDFs, reports, contracts, notes): Ethos 6 summarizes them, answers questions, and pulls out key facts, figures and action items.
• Images (photos, screenshots, charts, handwriting): Ethos 6 describes what it sees, reads the text in the image, and explains charts and diagrams.
• Videos (lectures, meetings, product demos): Ethos 6 summarizes what happens, finds key moments, and answers questions about the content.

A student can turn a recorded lecture into study notes. A founder can check a 50-page contract for key dates in minutes. A support team can read a customer's screenshot and spot the error straight away.

Atomesus Image 2: our own image generation model

Atomesus Image 2 is a separate model, built for one job: creating images. Describe what you want, and Image 2 turns your words into a finished visual in seconds.

"A minimalist product photo of a copper water bottle on a marble kitchen counter, soft morning light."

What you can create with Image 2:

• For creators: social posts, thumbnails, illustrations and concept art.
• For students: diagrams and visuals for projects and presentations.
• For businesses: product mock-ups, ad creatives and marketing images.

Because Image 2 is our own model, we control its cost and pass those savings on to you.

Better together

Ethos 6 and Image 2 work side by side in the same Atomesus workspace. For example, to create an ad for a new product:

1. Upload a photo of your product and ask Ethos 6 what makes it stand out.
2. Ask Ethos 6 to write three ad headlines and an image prompt.
3. Paste the prompt into Image 2 and generate your ad visual.

From idea to finished creative, without leaving Atomesus.

Availability

Both models are available today on atomesus.com.

• Ethos 6 is included in every paid plan, starting with Core at $1/month. Prime ($4/month) adds advanced reasoning and extended context.
• Image 2 generations are included in every paid plan. Plus ($10/month at launch, regularly $20) includes unlimited Image 2 generations under fair use.

See all plans at atomesus.com/pricing.

Ethos 6 API for developers

Developers can build with Ethos 6 using pay-as-you-go pricing, billed per use from a prepaid wallet:

• Input tokens: $0.52 per 1 million
• Output tokens: $2.06 per 1 million
• 1 million input + 1 million output tokens: $2.58
• One typical chat request (1,000 tokens in, 500 out): about $0.0016, or roughly 650 requests per $1

Getting started takes three steps in the Atomesus developer dashboard:

1. Activate the model API by creating or selecting Ethos 6 as the model to serve.
2. Create an API key and store it as ATOMESUS_API_KEY. Keys are shown only once, so save yours somewhere safe, and never put it in browser code.
3. Add wallet funds. Usage is billed per request, and charges appear in your dashboard after each completed request.

The API is OpenAI-compatible. If your app already uses the OpenAI SDK, you only need to change the base URL, API key and model name:

from openai import OpenAI
import os

client = OpenAI(base_url="https://apidev.atomesus.com/v1", api_key=os.environ["ATOMESUS_API_KEY"])
completion = client.chat.completions.create(
    model="ethos-6", messages=[{"role": "user", "content": "Hello!"}]
)
print(completion.choices[0].message.content)

Need streaming? Set stream: true to receive responses as Server-Sent Events. Examples for cURL, JavaScript, Go, Java, C#, PHP, Ruby and Rust are in the Atomesus API docs.

Try Ethos 6 and Image 2 today

Upload a document, ask about an image or describe the picture you want to create. Start now at atomesus.com.

We can't wait to see what you build, learn and create.

— The Atomesus Team