Skip to content

Is This the Cheapest Way to Use AI in 2026? (65+ Models Tested) | Iskander Malikov, Trioma

Eye On A.I.

55m
Transcript

Original podcast

Where to listen

Summary

Written by a model from the transcript; not yet reviewed.

This episode explores Trioma, a vertically integrated AI ecosystem that combines an orchestration engine running across 65 different AI models with a custom GPU network. The guest and the host discuss how shifting from traditional question-and-answer chatbots to agentic, result-oriented workflows is transforming software development, content creation, and consumer behavior. They explain how orchestration handles complex routing, subtasking, and bypassing model censorship, while the Trioma GPU Network allows users to acquire energy streams and swap them for AI credits to control token expenses.

  • Users want the end result rather than managing the reasoning process, which makes automated orchestration engines more practical than manual model selection. 4:24
  • By participating in the Trioma GPU Network, users can generate energy streams and swap them for AI credits or withdraw them. 13:39
  • Evaluating AI cost should be based on the total expense of achieving a satisfactory result rather than the raw price per million tokens. 23:52
  • Desktop applications provide AI agents with local machine autonomy and file access that secure web browsers cannot match. 35:14

Chapters

  1. 0:00
    Trioma’s Rapid Development and Core Concept

    Trioma built a vertically integrated AI ecosystem in ten minutes, emphasizing user-driven prompts and orchestration across 65 models.

  2. 0:33
    Trioma’s Layered Architecture and Model Orchestration

    Trioma’s three layers—interface, mid‑layer of 65 models, and GPU base—are explained, highlighting how the orchestration engine selects optimal models for tasks.

  3. 1:19
    Identity Management Challenges in AI Workloads

    The host discusses the rise of identity breaches and introduces Idira by Palo Alto Networks as a unified solution for human, machine, and AI identities.

  4. 4:24
    User‑First Approach: From Prompt to Result

    Trioma focuses on delivering end results rather than model processes, allowing users to simply state desired outcomes and letting the engine handle the rest.

  5. 10:14
    Token Economics and Enterprise Partnerships

    Trioma secures bulk discounts from model providers, balances token consumption, and offers a GPU network (TGN) that lets users earn or swap compute for credits.

  6. 16:00
    Competitive Edge: Orchestration vs. Single‑Model Platforms

    Trioma’s orchestration engine combines multiple models to avoid censorship and reduce iterations, offering a cheaper, more efficient alternative to competitors like Perplexity.

  7. 29:19
    Pricing Strategy and Transparency

    Trioma keeps token prices low, emphasizes total cost of achieving a result, and provides dashboards for users to track spend across models.

  8. 34:40
    Developer Experience: Desktop vs. Web

    Coders prefer Trioma’s desktop application for full local control, while the web platform offers convenience; Trioma supports both to cater to different workflows.

  9. 50:51
    Community‑Driven Growth and Global Reach

    Trioma prioritizes community engagement, education, and a global GPU network to democratize AI access and maintain sovereignty over models.

  10. 51:28
    Future Outlook: AI Sovereignty and Market Expansion

    Trioma plans to expand its GPU infrastructure, stay agile with new model releases, and focus on user education to navigate job displacement fears and empower creators.

  11. 54:52
    New Content Creation Trends

    The speaker highlights the surge of AI-generated series and content.

Highlights

  • 0:42

    Trioma Three Layer Architectureinsight

    The guest explains that Trioma is structured as a vertically integrated ecosystem with an interface layer, a mid-layer containing 65 models and an orchestration engine, and a bottom GPU compute layer.

  • 5:44

    Universal Access Strategystrategy

    Rather than targeting a narrow niche like coders or stay-at-home parents, Trioma aims for broad accessibility so everyone can leverage the universal knowledge stored in AI models.

  • 8:37

    Orchestration Bypasses Censorshipinsight

    The orchestration engine can automatically route a task to a different model if strict censorship on a private US-based model stops it from completing the request.

  • 13:39

    Trioma GPU Network Mechanicsstrategy

    The guest breaks down how customers buy software licenses in TGN to generate an energy stream, which can then be swapped for AI credits or withdrawn.

  • 23:52

    Cost of the Final Resultinsight

    The guest argues that raw token pricing becomes meaningless when an orchestration platform achieves a one-shot result that is ten times cheaper in total iterations.

  • 41:38

    Vibe Coding a Messenger Appstory

    The guest shares an anecdote about a co-founder who vibe coded a fully functional messenger app with audio, video, and chat support in just ten minutes.

  • 49:32

    Generating Long Video Contentdata

    The guest mentions that their studio platform can generate videos up to 60 minutes long, even though the underlying video model has a 30-second limitation.

Quotes

  • “Trioma did it in like 10 minutes.”
    Unnamed speaker · 0:00
  • “Non-human identities outnumber human identities 109 to 1,”
    Unnamed speaker · 1:44
  • “We live in a world where so many new applications will emerge.”
    Unnamed speaker · 0:33
  • “We have 65 different AI models, as well as the orchestration layer that connects them and makes them work together.”
    Unnamed speaker · 0:57
  • “We have a pretty hard working group of individuals, coders who've been building the product.”
    Unnamed speaker · 25:20
  • Idira by Palo Alto Networks 2:04–3:13

About this episode

What if you didn't have to choose between ChatGPT, Claude, and Gemini, and instead, the platform chose for you? That's the core idea behind Trioma, a vertically integrated AI ecosystem built by Iskander and his team in the UAE. Craig Smith sits down with Iskander Malikov, Co-founder & CEO of Trioma, to unpack how Trioma's orchestration engine routes prompts across 65 models, why token price is a misleading metric, and how their GPU network lets users participate in AI infrastructure build-out to offset their own compute costs. The conversation goes deep on the economics of AI subscriptions, the coming collapse of traditional SaaS, and why Trioma believes AI should be sovereign, accessible, and community-driven, not locked behind a single company's walled garden. If you're thinking about where AI access, cost, and infrastructure are heading, this one is essential listening. Subscribe to Eye on A. I. for weekly conversations with the people building and deploying the future of AI.

Transcript

0:00
55:13

9,043 words · 2 speakers

Topics

  • Trioma
  • Perplexity
  • Frontier AI Labs
  • Google
  • Anthropic
  • GLM
  • Palo Alto Networks
  • OpenAI
  • Qwen
  • HeyGen

Page updated · transcript and summary are machine-generated