Sign in

Balanced · Multimodal

Production chat, coding, vision + audio

Balanced · Multimodal

Meet Yuva

Best for production chat, coding, vision + audio.

Yuva is the production default and where most work should sit. It reads images and audio as well as text, generates images and audio, and calls several tools at once rather than one after another. Its 1M context is large enough to hold a repository, which is what makes it the coding model — it can be shown the code rather than told about it. Up to two stages, so it considers before it answers without the cost of a full agentic run.

Balanced · MultimodalModels

Listen to article

In short

  • The production default — where most work should sit
  • Reads images and audio as well as text, and generates images and audio
  • Calls several tools at once rather than one after another
  • A 1M-token context, large enough to hold a repository
  • Considers before it answers, in up to two stages
0:13 / 0:17

Coding against the real repository

Yuva's 1M context is large enough to hold a repository, which is what makes it the coding model: it can be shown the code rather than told about it.

It works in up to two stages, so it considers a change before it answers without the cost of a full agentic run.

Images and recordings in the conversation

Anything with an image or a recording in it belongs on Yuva. It reads screenshots, photos and audio, and it can generate images and audio in return.

Several tools in one turn

Yuva calls tools side by side instead of one after another, which suits production chat and work that has to gather from more than one place before it can answer.

Keep your spreadsheet up to date by talking about the work

Turn everyday updates into changes in your connected Google Sheet. Explain what changed, review the proposed edits, and keep working in the same conversation.

Make invoice review easier to work through

Compare invoices with your records using rules you define. Get a clear exceptions report so your review starts with the items that need attention.

1 / 10

Make the portal part of the workflow

Use Upfyn with a supported browser to work through repetitive portal tasks. Provide the source information, review the plan, and check the result.

Get the idea out of your head and onto a canvas

Create decks, documents, designs and websites in Upfyn Creative. Explain the purpose, develop a draft, and refine it through chat on your design system.

Build the small tool your everyday work keeps asking for

Describe a repetitive task in plain language and develop a small tool with Upfyn. Keep the project, review changes, and test the result with real examples.

Where it runs

Yuva is available wherever you use Upfyn: the desktop app and your phone. Your projects, tools and permissions stay with you when you switch between Bala, Yuva and Rishi.

How the three compare

BalaYuvaRishi
Best forEveryday chat, drafting, high-volume textProduction chat, coding, vision + audioHard multi-step reasoning, research, media
SpeedFastestFastDeliberate
DepthDirect — single passConsidered — up to 2 stagesAgentic — most thorough
ModalityTextText + image + audioText + image + audio + video generation
Contextup to 1M1M1M
PlansEvery plan, including FreeEvery plan, including FreeEvery plan · usage charged in credits

Where it stops

No video input — that is Rishi. Costs more per token than Bala, so high-volume text work is cheaper on Bala unless it needs the extra care.

Availability

Every plan, including Free. Usage counts toward your plan’s limit; the plans and what each includes are on the pricing page.

See pricingTry it in Upfyn

Specifications

Speed
Fast
Depth
Considered — up to 2 stages
Modality
Text + image + audio
Context
1M (1,000,000 tokens)
Plans
Every plan, including Free

The other two

Fast · EverydayBala Everyday chat, drafting, high-volume textDeep · AgenticRishi Hard multi-step reasoning, research, media

Keep reading