Barathkanna

6 ms·

1.
by Barathkanna 6mo ago
Sounds like a plan, But what if you can just pay a fixed cost every month and not worry about anything?
2.
by Barathkanna 6mo ago
That’s true, but AI is interesting because consumption-based pricing introduces a lot more variance than typical SaaS infrastructure. One user action can trigger dozens of model calls in an agent workflow. That’s partly why we started exper
3.
by Barathkanna 6mo ago
Local models help remove token cost uncertainty, but they shift the problem to infrastructure and ops. GPUs, scaling, maintenance, and latency can add up quickly depending on the workload. For many builders it ends up being a tradeoff betwe
4.
by Barathkanna 6mo ago
That’s great. Real-time tracking is a big step already. The tricky part we kept running into was the variance itself, especially with retries and agent loops. That’s partly why we started experimenting with Oxlo.ai ( https://oxlo.
5.
by Barathkanna 6mo ago
One underlooked source of variance is retries from formatting failures. In many agent systems the loops dominate the cost, not the raw token length. We ran into the same issue building agent workflows, which is why we started building http
6.
by Barathkanna 6mo ago
Agreed. The real cost unit becomes the whole agent workflow, not a single LLM call. One user action can trigger dozens of calls. We ran into the same issue and ended up building https://oxlo.ai to make the cost side more predict
7.
by Barathkanna 6mo ago
Exactly. That’s actually why we started building Oxlo.ai. Early stage builders usually just want to experiment without worrying too much about token cost spikes.
8.
by Barathkanna 6mo ago
True, but for early stage builders it’s harder to design those guardrails upfront. A lot of the time you only discover the retry patterns and cost spikes once real users start hitting the system.
9.
by Barathkanna 6mo ago
Local models solve the marginal cost problem, but they move the complexity into infrastructure and throughput planning instead.
10.
Ask HN: How are people forecasting AI API costs for agent workflows?
5 points by Barathkanna 6mo ago|22 comments
11.
Show HN: Oxlo.ai – AI APIs with unlimited tokens and request based pricing(oxlo.ai)
1 points by Barathkanna 7mo ago|1 comments
12.
by Barathkanna 8mo ago
Agreed. Self-hosting gives the cleanest fixed cost, but you pay for it in ops and capacity planning. I’m mainly curious whether there’s a middle ground that gives early teams more predictable spend without immediately taking on full infra o
13.
Ask HN: How do you budget for token based AI APIs?
1 points by Barathkanna 8mo ago|4 comments
14.
by Barathkanna 8mo ago
I asked GPT for a rough estimate to benchmark prompt prefill on an 8,192 token input. • 16× H100: 8,192 / (20k to 80k tokens/sec) ≈ 0.10 to 0.41s • 2× Mac Studio (M3 Max): 8,192 / (150 to 700 tokens/sec) ≈ 12 to 55s Th
15.
by Barathkanna 8mo ago
That won’t realistically work for this model. Even with only ~32B active params, a 1T-scale MoE still needs the full expert set available for fast routing, which means hundreds of GB to TBs of weights resident. Mac Studios don’t share unifi
16.
by Barathkanna 8mo ago
A realistic setup for this would be a 16× H100 80GB with NVLink. That comfortably handles the active 32B experts plus KV cache without extreme quantization. Cost-wise we are looking at roughly $500k–$700k upfront or $40–60/hr on-demand
17.
by Barathkanna 8mo ago
TLDR: AI didn’t diagnose anything, it turned years of messy health data into clear trends. That helped the author ask better questions and have a more useful conversation with their doctor, which is the real value here.
18.
by Barathkanna 8mo ago
TLDR: IPv4 is fully exhausted and no longer growing. Internet growth now depends on IPv6 adoption and address sharing, but IPv6 rollout is still uneven across regions.
19.
by Barathkanna 8mo ago
I get why this exists and appreciate the transparency, but it still feels like a slippery middle ground. Age prediction avoids hard ID checks, which is good for privacy, yet it also normalizes behavioral inference about users that can be wr
20.
by Barathkanna 8mo ago
What’s interesting here isn’t the humanoid form factor, it’s the systems integration. Plugging robots into Siemens’ industrial stack means they’re being treated like first-class nodes in existing logistics workflows, not special demos. If h
21.
by Barathkanna 8mo ago
TLDR: Soft deletes look easy, but they spread complexity everywhere. Actually deleting data and archiving it separately often keeps databases simpler, faster, and easier to maintain.
22.
by Barathkanna 8mo ago
Thank you!! We are definitely fully focused on Developer experience. Would love some feedback if it looks interesting
23.
by Barathkanna 8mo ago
Totally fair question, and you’re not being negative. We’re not claiming better token economics in the sense of magically cheaper tokens, and we’re not just burning money to subsidize usage indefinitely. You’re right that this isn’t a new p
24.
by Barathkanna 8mo ago
That’s fair, and I probably didn’t explain it clearly. We’re building an AI API as a service platform aimed at early developers and small teams who want to integrate AI without constantly thinking about tokens at all. I agree that token eco
25.
Ask HN: Is token-based pricing making AI harder to use in production?
3 points by Barathkanna 8mo ago|7 comments
26.
by Barathkanna 8mo ago
I agree with the core concern, but I think the right model is smaller, not zero. One or two strong technical writers using AI as a leverage tool can easily outperform a large writing team or pure AI output. The value is still in judgment, c
27.
by Barathkanna 8mo ago
This mostly changes how location is requested, not what you can do with it. Instead of imperative JS calls, location access becomes declarative in HTML, which gives browsers more context for permission UX and auditing. Your app logic, data
28.
by Barathkanna 8mo ago
This is pretty eye-opening. It really drives home how simple the core control logic can be. Starting with toy cars or small-scale vehicles feels like a great way to teach and validate these ideas before layering on unnecessary complexity.
29.
by Barathkanna 8mo ago
As an edge computing enthusiast, this feels like a meaningful leap for the Raspberry Pi ecosystem. Having a low-power inference accelerator baked into the platform opens up a lot of practical local AI use cases without dragging in the cloud
30.
by Barathkanna 8mo ago
For those wondering how this differs from Nvidia GPUs: Nvidia = flexible, general-purpose GPUs that excel at training and mixed workloads. Furiosa = purpose-built inference ASICs that trade flexibility for much better cost, power efficiency

More ›