Thomas Yu

eng @FireworksAI_HQ // ex elevenlabs robinhood

[Thomas Yu @tommyyu]

@MiaAI_lab Would love to add Fireworks serverless integration into this once its open source

[Thomas Yu @tommyyu]

@marvinvonhagen @cognition I was already pushing out a ton of code 😅

But I’m sure it’ll scale up as I get more acquainted

[Thomas Yu @tommyyu]

I've been using @cognition's new [REDACTED] product over the last 24 hours and wow

I've tripled my productivity overnight

Very impressed so far

[Fireworks @FireworksAI_HQ]

We evaluated Nexus on 7 benchmarks and 1,222 tasks. It defines the Pareto frontier.

Mix open and closed models. Get better intelligence for 4x lower cost than a single frontier model.

The smartest way to extend your Claude, ChatGPT, or Codex plan, starting at $8/mo.

https://t.co/pwCYfD3OhS

[Thomas Yu @tommyyu]

@dwehner Welcome!

[Thomas Yu @tommyyu]

@DesignArena @thinkymachines @Kimi_Moonshot incredible work

[Thomas Yu @tommyyu]

@APompliano @cfosilvia What models are you using over Claude right now?

[Thomas Yu @tommyyu]

@zachmoskow @FireworksAI_HQ should be fixed now

[Thomas Yu @tommyyu]

@anoushkavaswani And it's only getting better

[Thomas Yu @tommyyu]

@zachmoskow @FireworksAI_HQ taking a look, will get this fixed

[Fireworks @FireworksAI_HQ]

Meet the new Nexus.

Nexus helps individuals and enterprises use open and closed models in their existing AI tools, choosing the right model and helping lower the cost of each task.

Getting started is simple with subscriptions from $8/month: https://t.co/4e9ZagqbSB

[Thomas Yu @tommyyu]

@FireworksAI_HQ You can sign up for Nexus here: https://t.co/TuQqJpqYDJ

[Thomas Yu @tommyyu]

Today @FireworksAI_HQ is launching Nexus, our new model router, to the public.

Nexus can achieve better than frontier performance by using the "best" model for different tasks. Different models have their strengths and weaknesses - Nexus uses this to smartly route tasks.

Nexus reduces token spend without compromising on performance by leverage task and cost awareness for every turn.

We are live today - our CLI and desktop app make it incredibly easy for users to start saving on AI spend in just a few clicks.

[Vals AI @ValsAI]

Nexus Auto (Sol) by @FireworksAI_HQ 's is the first model to reach 60% on the Vals Router Index, now ranking #1 at just $1.25/test.

It leads the previous #1, Router (Ramp), by roughly 2 points.

[Thomas Yu @tommyyu]

@ValsAI @FireworksAI_HQ Still have more improvements coming!

[Thomas Yu @tommyyu]

@rikiyatakehi @a1zhang @lateinteraction @jayden_teoh_ Really cool research done here

[Thomas Yu @tommyyu]

Personal update: I have joined @FireworksAI_HQ to work on intelligent model routing on the Nexus team!

I'm excited to help build a cohesive tool that can deliver frontier level intelligence at a fraction of the cost by leveraging both open and closed models.

More to come soon!

[Thomas Yu @tommyyu]

@felixrieseberg @trq212 Other than system design questions and take home assignments, what else do you find effective for measuring engineering excellence during interviews?

[Thomas Yu @tommyyu]

@amorriscode You're a legend, thanks for all the fantastic work on Claude Code desktop

[Fireworks @FireworksAI_HQ]

We have another release you don’t want to miss.

@MiniMax_AI M3.1 Flash is now live on Fireworks, day 0 with the open-weights release.

Built for agents that run long, multi-step work:

→ Cowork automation: 62.8 on AutomationBench-AA
→ Coding: 73.7% on DeepSWE v1.1
→ Visual reasoning: 91 on BabyVision
→ $0.15 input / $0.60 output / $0.015 cached per 1M tokens, half the price of M3

We've heard your feedback that you want more models on serverless. So we're launching more models in preview mode, starting today with M3.1 Flash, available for two weeks.

We'll be watching demand, and if the community likes it, we'll keep it around.

Try it now → https://t.co/RVwd3ETz9d