Sabitlenmiş Tweet
Token-maxxing is getting out of hand.
Most AI apps send every request to the biggest model, even when a smaller model would work.
We built Dari Router to fix that.
English
Avyay Varadarajan
59 posts
Avyay Varadarajan
@avyvar
co-founder @daridotdev // prev @ycombinator, @caltech, basketball data @lakers, software @uber, startups
Fremont, CA Katılım Nisan 2018
563 Takip Edilen564 Takipçiler
Grok 4.5 is out! 🚀
I spent the past few months as part of the Coding RL Bigrun and RL Science team, pushing RL scaling to help deliver Grok 4.5.
Proud to have contributed to this model, working on core RL algorithm changes and executing bigruns at true frontier scale with the team.
Huge shoutout to everyone at @SpaceXAI who helped make this happen!
Artificial Analysis@ArtificialAnlys
SpaceXAI’s Grok 4.5 scores 54 to place fourth on the Artificial Analysis Intelligence Index following only Fable 5, GPT-5.5, and Opus 4.8. It scores on par with GPT-5.5 in Codex on the Artificial Analysis Coding Agent Index in the Grok Build harness, at much lower cost Grok 4.5 improves 16 points over Grok 4.3 on the Intelligence Index, bringing SpaceXAI to the intelligence frontier behind only OpenAI and Anthropic, and outperforming all open weights models and notably Google’s Gemini models. Key standout areas of performance are agentic knowledge work and coding. Grok 4.5 in Grok Build scores 76 on the Artificial Analysis Coding Agent Index, on par with GPT-5.5 (xhigh) in Codex and just below Fable 5 (max) in Claude Code, and at a small fraction of the token usage and price. Congratulations to @SpaceXAI, @cursor_ai, and @elonmusk on the impressive release! Key Takeaways: ➤ Grok 4.5 performs very strongly on agentic tasks. Grok 4.5 ranks #4 on GDPval-AA v2 with an Elo of 1543, between Claude Opus 4.8 (1600) and GLM-5.2 (1513). It achieves the top score on 𝜏³-Banking of 33%, above 31% from GPT-5.5 (xhigh), and sits on the cost vs performance Pareto frontier across all three agentic evaluations in the Intelligence Index ➤ Grok 4.5 is one of the most cost efficient models to run for near-frontier intelligence. It costs $0.31 per task on the Artificial Analysis Intelligence Index and $2.59 per task on the Artificial Analysis Coding Agent Index within Grok Build ➤ Low cost for Grok 4.5 is driven by both low pricing and token efficiency. Grok 4.5 has a headline price over 60% lower than Claude Opus 4.8 and GPT-5.5, and used ~14k output tokens per Intelligence Index Task - over 60% lower than Opus 4.8. On the Coding Agent Index, Grok 4.5 stands out on the Pareto frontier of Coding Agent Index score vs. Total Tokens, using only 1.9M tokens for the Coding Agent Index while scoring 76 ➤ As a coding agent, Grok 4.5 in Grok Build is on par with GPT-5.5 and offers efficiency benefits: In our Artificial Intelligence Coding Agent Index that consists of DeepSWE, Terminal-Bench v2, and SWE-Atlas QnA, Grok 4.5 in Grok Build ranks third, on par with GPT-5.5 (Codex) and below Fable 5 (Claude Code). It is also very efficient in achieving this result: Grok 4.5 in Grok Build cost $2.49 per task while Fable 5 in Claude Code cost $11.80 and GPT-5.5 in Codex $5.07. This is driven by relatively low token pricing and the model using far fewer tokens than comparable models (1.9M average tokens used per task), significantly less than Fable 5 in Claude Code (7.2M) and GPT-5.5 in Codex (6.2M) Other model details: ➤ Context window of 500k tokens - a reduction from Grok 4.3’s 1M token context, but retaining configurable reasoning and vision input ➤ Pricing of $2/$6 per 1M tokens of input/output; cache hits are discounted by 75% to $0.5 per 1M tokens, and costs still double with long (>200k token) inputs ➤ As Elon Musk has disclosed, Grok 4.5 is 3x larger than its predecessor at 1.5T parameters
English
(player's association nightmare, but interesting to think about)
English
another reason to feel like NBA teams / sports orgs are undervalued:
teams capture every player's 3d skeletal data (hawkeye data) at ~25hz in games / practice. if players wear eye-tracking goggles (there are real coaching incentives to this) you get real-world egocentric data
Reka@RekaAILabs
Training world models needs egocentric video and dense action signals, synchronized. That data is genuinely hard to find. We built it from Counter-Strike 2 demos. CS2-10k: 600K+ player-round videos, 10K+ hours, per-frame annotations — keyboard state, mouse delta, 3D position, camera yaw/pitch. All paired to the visual stream. Why CS2 demos? Matches are deterministic replays. We can reconstruct first-person video and extract the exact control inputs that caused every visual change. No labeling, no estimation. We're also releasing cs2-dem-renderer — the open-source pipeline we used to build it. Give it a .dem file, it outputs .mp4 + .parquet. Run your own dataset at whatever scale you need.
English
@Walter64456133 If you use the SLM option, you do have the added overhead.
However, if you use the heuristic router (a combination of input token + eval scores), there’s negligible latency.
English
@avyvar But isn't that it reduces the speed of response we get through the api.
English
Token-maxxing is getting out of hand.
Most AI apps send every request to the biggest model, even when a smaller model would work.
We built Dari Router to fix that.
English
@halfdebugged this depends on your router of choice
If you use the SLM option, you do have the added overhead.
However, if you use the heuristic router (a combination of input token + eval scores), there’s negligible latency.
rn, everything is from the global endpoints
English
If you're a company struggling with token costs and want us to take a look, book a call.
calendly.com/avyay-dari/30m…
English
@daridotdev gives you full control over your routing workflows.
Use your own traces and evals to build custom routers for each workflow: fine-tuned small-model routers when you need learned behavior, or deterministic policies when you want full control.
English
tons of interesting benchmarks you could do within a front office since you have counterfactuals
presti-bench: given all the internal data we have on players / how we value draft picks, would you accept this trade proposal?
English
there are several interesting basketball benchmarks one could do but would be cool to see how well LLMs perform at predicting rotation charts
imperfect information, but given recent stats, lineup data, etc, how close can an LLM emulate coach game-planning
Theo - t3.gg@theo
We need more niche benches. We need ios-bench. We need ts-bench. We need baseball-bench. We need yt-thumbnail-bench. We need way more creativity in how we measure what models can do.
English
Everyone seems to only focus on sandboxes for agent runtimes
but decoupling your agent loop from the sandbox / persisting state via durable execution is tablestakes for long-running agents, as seen w/ Cursor
if you want to scale long-running agents try us out @daridotdev!
Cursor@cursor_ai
A great cloud agent experience involves a lot more than moving a local agent to a server. We've learned that it requires a durable execution platform, a powerful harness, and the tools and infra to give agents realistic development environments. cursor.com/blog/cloud-age…
English
Built a tool for sharing images/files to my remote agents: contextdrop.dev
Upload a file and get a temp link or login on both machines and let agents pull files.
I liked the simplicity of @bentlegen’s glance.sh but wanted something similar as a pure CLI.
English
@Xylon_lew @daridotdev on every new message sent we try to reconnect to the sandbox as if its warm - if e2b has unexpected behavior ie pausing sandboxes, we fallback to resuming the sandbox. if the sandbox is dead, then we rehydrate via the durable session/checkpoint.
English
We made deploying agents to prod so easy your coding agent can do it in under 5 minutes.
Get a production-grade API that scales to 100s of concurrent sessions, pauses for days between messages, and recovers sessions even on sandbox crashes or LLM provider failure.
@daridotdev
English
Get started with $10 of free credits at app.dari.dev or book a demo at calendly.com/avyay-dari.
English
"none of github's features work reliably anymore, everyone's gonna churn"
github is not a car
English
Avyay Varadarajan retweetledi
Agent infra is undifferentiated but hard to do well, delegate to this cracked team instead and focus on your vertical’s differentiation🧠
Avyay Varadarajan@avyvar
Introducing @daridotdev. Deploy your AI agent as a stateful, versioned API. - durable sessions - built-in execution + sandboxes - tools, files, retries handled Stop stitching infra. Just dari deploy.
English