Idle memory in the TUI: Loop: 17 MiB
Pi: 140 MiB
Claude Code: 240 MiB
opencode: 761 MiB Your harness shouldn’t use more RAM than your agent’s workload does.

It does the same work with far less. Loop used 803k tokens against OpenCode’s 1.82M and Goose’s 1.33M. That works out to $0.13 vs $0.29 for the full suite, and it finished in 12 minutes vs 23.

We benchmarked 4 harnesses on identical coding tasks (website, JS, Rust), all running GLM 5.3 Flash and graded them. Loop: 13/13 passed
Fastest median task time: 29s (OpenCode: 72s)
Lowest CPU time: 1.7s (OpenCode: 8.5s)

We just open-sourced Loop: a Rust agent harness that’s 44x lighter than opencode at idle, 5x lighter on CPU, and finishes the same tasks in half the time. Initial benchmarks look promising. Same model. Same tasks. 13/13 passed. Developer preview, out now. 🧵 > curl -fsSL loop.soket.ai/install | bash github.com/soketlabs/loop

Horrible experience with @wandb. Have been trying to upgrade for weeks now and all 9 cards are declined. Seems like they dont support India for billing. Lost access to the account and the support takes 3 days to response even with priority support. Moving back to tensorboard. If an Indian entity is working on Wandb alternative would love to get a subscription. Also, India lets please break this US dominance on all essential softwares.

I think low frequency tokens are replaced by high frequency indic sub-words. Identical token IDs and placement is only possible if the training data distribution is the same as Gemma which I dont think is the case.

BTW models developed under Project EKA will be open source. More on the initiative here: soket.ai/project-eka

Many youngsters are enthusiastically finding vulnerabilities in govt websites. @IndianCERT should use this opportunity to launch a bug bounty program to incentivise and encourage them. Will build trust and also safeguard the Indian infra. Wont find a better opportunity than this.

Here’s an early signal to what I said For almost an year now, we at @soketlabs have been working on curating a frontier scale pretraining data corpus along with finding the best architecture that fits the diversity of languages along with being compute optimal for both training and inference. Sharing one of the many successes we have encountered. Our current version of the arch (yeah, its not a clone of deepseek or any other known arch) is at least 30% compute optimal to Deepseek’s sparse-MoE. These are just initial results and we hope to find a lot more. Shows that we have a lot more to learn about these architectures. Also excited about the pretraining data we have curated but more on that later Research efforts take time but they also yield exponential outcomes and thats what most people in India should be building towards

So majority power points at T1 IGI don’t work. Its like a puzzle to find the one that does. Looks like no one is maintaining the infra. @DelhiAirport please take a note

Interesting conversations happening on AI in Defence, sovereign capabilities in Infra, why an Indian model. Impact Summit, Meeting Room 6, Bharat Mandapam