@avyvar
Account based inUnited States
About this account
- Account based in
- United States
- Connected via
- United States App Store
Account-level information from X, not a live location or the device used for a specific post.
co-founder @daridotdev // prev @ycombinator, @caltech, basketball data @lakers, software @uber, startups
Fremont, CA
Joined April 2018
- Tweets89
- Following653
- Followers677
- Likes4.2K
Pinned Tweet
Today, we're releasing our open-weight, auto-routing model @daridotdev, built for coding agents.
We're state-of-the-art on the Pareto Frontier, w/ 70% cost reduction + comparable coding performance to Fable.
Bring your own evals, choose your models, or use our defaults.
Avyay Varadarajan retweeted
Today we’re launching Keythorn.
AI isn’t just thinking anymore. It’s moving money, talking to customers, and taking action. Every new agent brings new risk.
The future runs on AI. We’re building the insurance for it.
Meet @TryKeythorn ↓
Today we're launching
> the world's fastest TTS endpoint at 50 ms time-to-first-audio (TTFA)
> the world's cheapest modern TTS endpoint at $5 / 1M characters
Powered by Qwen3-TTS 1.7B on Nari Labs inference engine.
🚀 5x faster than Cartesia
💸 10x cheaper than ElevenLabs
🎙️ expressive voices, outperforming industry average
🎁 available FREE for a limited time
We believe open-source will win: not just in LLMs, but also in multimodal AI. Nari Labs is here to accelerate that future, starting with speech.
Try now at narilabs.com
Avyay Varadarajan retweeted
A 14B open model on a single RTX 4090 matched our hosted frontier model on text-to-SQL.
That wasn’t supposed to happen.
We ran 28 configurations, graded 25,000+ answers, and benchmarked against Snowflake + Databricks.
What we learned:
model size, model reputation, and vendor benchmark scores are bad shortcuts for predicting performance on your database.
So today we’re open-sourcing mnemiq.
Test it on your own data ↓
Avyay Varadarajan retweeted
token pricing is effectively meaningless now. 3.8 flash looks 13x cheaper when measured per token, but Astra is cheaper per task since it's far more efficient. measure your costs per task, not per token.
if you're running out of Fable weekly limits try running
`dari --claude`
Use Fable + open models and make your subscriptions last for 2x longer with no quality degredation :)
Literally unusable. The rate limits are absurd. Oh, and by the way, Fable’s automatic continuation is bugged and doesn’t even work.
I honestly don’t know why I still bother using Claude at this point. 5.6 is simply better overall anyway.
Give me GPT-Astra and im fine. its so frustrating. seriously.
oh, and btw. For subscription users, Anthropic has not announced lower prices or higher usage limits regarding Fable 5.1s efficency gains; the savings explicitly apply “wherever usage is billed by token,” so greater efficiency within Pro or Max subscriptions possible not gonna happen.
some thoughts re: subagents
subagents are *great* if you’re selling tokens. increasing token throughput is a massive win for a token seller.
nobody has the bandwidth to manage 100 agents concurrently, but if all it takes is one prompt and “ultracode”, you can easily 10x usage!
of course, the appeal of multi-/sub-agent is massive. you essentially get infinite context if done right
(the hugging face exploit is clear evidence of the potential/danger of multi-agent)
the labs will obviously continue with multi-agent RL and performance will only improve.
Avyay Varadarajan retweeted
Overall, I’m very surprised at how little media coverage there’s been around the OpenAI / Hugging Face attack. It’s clearly one of the most important things to happen this year.
Model routers and evals go hand-in-hand - router decisions should be based on evals, and you need to eval different models/harnesses within the exact (or very similar) environments that your agents act in to properly measure savings
this is why we built self-bench! (github.com/mupt-ai/self-benc…) - it's fully open-source and lets you build simple @harborframework evals from your PRs
Public evals are saturated. What matters is that a model reliably works in your codebase.
To solve this, we built self-bench (github.com/mupt-ai/self-benc…) - an open source package to automatically build evals on your own private PRs.
One CLI command -> get a clean Harbor dataset.
Companies are trying to reduce token costs by using smaller / open-source models, but switching without evals is flying blind.
We built evals based on some popular OSS packages like Next.js and Vite. @daridotdev 's auto-router is on the pareto-frontier in all of our tests!
If you want to reduce your token costs without sacrificing performance, you can now:
- build an eval on your repo
- run open weight models, frontier models, and our router
- see cost/performance/latency tradeoffs!
book a call to reduce token cost: calendly.com/avyay-dari/30mi…
Avyay Varadarajan retweeted
We had a great time at last week's Agents & APIs SF Developer Meetup! Thanks to @kwindla of @pipecat_ai, @codemathics of @coderabbitai, Matt Colozzo of @astropods_ and @avyvar of @daridotdev for presenting! 🫶
To get here, we finetuned an SLM to be benchmark, cost, and cache aware.
Using our SLM adds negligible latency - we only use the results for the next query in an agent loop.
Blog: dari.dev/blog/auto-routing-m…, dari.dev/cache-cost-curve/in…
Model link: huggingface.co/dari-ai/route…
Get started at dari.dev
If you are an enterprise looking to reduce and/or understand their LLM spend - book a meeting.