• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
AI
Announcement

River API is pitched as a cheap route to task-specific open models

Igor Babuschkin claims fine-tuned open-weight models can beat GPT-6 Astra and Claude Opus 5.5 for under 1% of the cost.

clem 🤗C🤗
Sriram KrishnanSK
Igor BabuschkinIB
8 Sources, 2d ago, first seen 2d ago

TLDR

Igor Babuschkin is promoting River API as a way to fine-tune open-weight models for specific tasks. He claims the resulting models can outperform GPT-6 Astra and Claude Opus 5.5 for less than 1% of the cost, but the posts provide no benchmark methods, task list or pricing calculation. Babuschkin also predicts customized open models will handle most global LLM usage within a few years.

Combined views

48K

8 Sources, first seen 2d ago

459 likes23 comments294 saves77 reposts

Combined views

48K

8 Sources, first seen 2d ago

459 likes23 comments294 saves77 reposts

In a promotional post, Igor Babuschkin says River API can fine-tune open-weight models to outperform GPT-6 Astra and Claude Opus 5.5 on tasks selected by the user. He claims the approach requires less than 1% of the cost of the named models.

Featured Source

The comparison is task-specific, not a claim that one fine-tuned model is generally stronger. The posts do not include benchmark methods, a task list, model recipes or a calculation behind the cost figure, so the performance and price claims cannot be independently evaluated from the available evidence.

Babuschkin also predicts that customized open-weight models will handle most LLM usage worldwide within a few years as building them becomes easier and cheaper. The replies are enthusiastic but do not validate that forecast or the benchmark claim: Eric Zelikman called it very cool work, Anjney Midha called it incredible, and Jeff Huber pointed to the task-to-synthetic-data-generation step as a bottleneck.

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Related Videos

  • Hemant Taneja with River AI's Igor Babuschkin: Own Your AI ExperienceGeneral Catalyst · YouTube
  • xAI Co-Founder: Igor Babuschkin Unpacks the Future of Model DevelopmentUnsupervised Learning: With Jacob Effron · YouTube

Useful links

River AI

Introducing River AI

General Catalyst · YouTube

Hemant Taneja with River AI's Igor Babuschkin: Own Your AI Experience

RuntimeWire

Igor Babuschkin's River AI raises $1.1 billion for customer-owned models

Unsupervised Learning: With Jacob Effron · YouTube

xAI Co-Founder: Igor Babuschkin Unpacks the Future of Model Development

River AI

Blog — River AI

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Useful Links

River AI

Introducing River AI

RuntimeWire

Igor Babuschkin's River AI raises $1.1 billion for customer-owned models

River AI

Blog — River AI

Related Videos

  • Hemant Taneja with River AI's Igor Babuschkin: Own Your AI ExperienceGeneral Catalyst · YouTube
  • xAI Co-Founder: Igor Babuschkin Unpacks the Future of Model DevelopmentUnsupervised Learning: With Jacob Effron · YouTube

Useful Links

River AI

Introducing River AI

RuntimeWire

Igor Babuschkin's River AI raises $1.1 billion for customer-owned models

River AI

Blog — River AI

Related

River AI Raises $1.1 Billion for User-Owned AI

Multiple AI researchers reply with congratulations to Igor Babuschkin.

Grok Bot reportedly shifting to a best-model router, with Claude Opus 5.5, MidJourney and Suno in the mix

Elon Musk said Grok Bot will use “the best back end model for any given task,” naming Claude Opus 5.5, MidJourney, Suno and other APIs in a future-facing plan rather than a confirmed live rollout.

Claude Opus 5.5 reportedly scores 94.2% on NerfBench, within normal variance

BridgeMind AI says Opus 5.5 was scoring above its launch level a day earlier but cautions that the dip is not yet evidence of a nerf.

8 Sources

Igor Babuschkin@ibabWe show how you can use the River API to fine-tune open weight models to beat GPT-6 Astra and Claude Opus 5.5 on tasks you care about, with less than 1% of the cost! A few years from now most of the LLM usage in the world will run on customized open weight models. Building your own custom models is getting easier and more affordable. Our mission is to help everyone own their own intelligence and be independent of big AI labs.2d
Eric Zelikman@ericzelikmanVery cool work!2d
Jeff Huber@jeffreyhuber@ibab make task -> SDG easy! thats the bottleneck...2d
Anjney Midha@AnjneyMidhaincredible2d
Beff (e/acc)@beffjezos@AnjneyMidha Decentralization in the space of weights is coming2d
Sriram Krishnan@sriramk@ibab Great work.1d
clem 🤗@ClementDelangueRT @ibab: We show how you can use the River API to fine-tune open weight models to beat GPT-6 Astra and Claude Opus 5.5 on tasks you care a…1d
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    River AIRiver APIIgor Babuschkin
    Claude Opus 5.5

    8 Sources

    Igor Babuschkin@ibabWe show how you can use the River API to fine-tune open weight models to beat GPT-6 Astra and Claude Opus 5.5 on tasks you care about, with less than 1% of the cost! A few years from now most of the LLM usage in the world will run on customized open weight models. Building your own custom models is getting easier and more affordable. Our mission is to help everyone own their own intelligence and be independent of big AI labs.2d
    Eric Zelikman@ericzelikmanVery cool work!2d
    Jeff Huber@jeffreyhuber@ibab make task -> SDG easy! thats the bottleneck...2d
    Anjney Midha@AnjneyMidhaincredible2d
    Beff (e/acc)@beffjezos@AnjneyMidha Decentralization in the space of weights is coming2d
    Sriram Krishnan@sriramk@ibab Great work.1d
    clem 🤗@ClementDelangueRT @ibab: We show how you can use the River API to fine-tune open weight models to beat GPT-6 Astra and Claude Opus 5.5 on tasks you care a…1d
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet