Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Sakana Fugu

sakana.ai

Post reply on HN
Print view View original

101–110 of 141 posts

Re: Sakana Fugu

#101

Post by erispoe »

mark_l_watson wrote: post #93
holistio wrote: post #26

Earlier quoted context omitted.

Maybe. But for now it's fascinating how $200/month has kind of become a normal tier. It's similar to how AirPods normalised all of us having $300+ headphones. All of us would have scoffed at the idea a decade ago.

But, it is not all about cost: models like DeepSeek v4 flash (I use the US company Fireworks.ai and also buy tokens directly from DeepSeek) is very fast, very low latency while working. Would you want to use a text editor that updates the screen very slowly? Kind of the same thing for using agentic systems as coding assistants: don’t want a ‘sluggish’ experience.

I have, mostly, long running autonomous tasks, so it doesn't matter how slow inference is. If I optimize for latency it means I'm turning into the limiting factor.

Re: Sakana Fugu

#104

Post by bicx »

holistio wrote: post #23

You pay $200/month to Anthropic, $200/month to OpenAI, $200/month to Cursor, $200/month to $200/month to Google, and seeing that it didn't come to a nice round $1024/month, you pay $200/month to Sakana to coordinate it all, because why not. While you're at it, feel free to send me $200 as well, I'll generate a crypto address ending with "AI".

I wish I only paid $200/mo for Anthropic! Multiply that by 20x.

Re: Sakana Fugu

#105

Post by andai »

See also: OpenRouter Fusion, similar idea, although it seems limited to internet research tasks? (Unclear, maybe someone who has used it can elaborate.)

What's nice is that OpenRouter included a pareto graph showing the cost as well as the performance. (But not time, unfortunately -- model fusion adds a large factor to round trip time.) Benchmarks are a lot less helpful without that.

OpenRouter: Surpassing frontier performance with fusion (blog post with benchmarks)

https://news.ycombinator.com/item?id=48525392

OpenRouter Fusion API

https://news.ycombinator.com/item?id=48537641

See also: Sibling comment with an open source implementation

https://news.ycombinator.com/item?id=48624782#48629598

I did my own last weekend in a few lines of Python, though I haven't tested it much yet. (Looking for some very hard, very cheap benchmarks, if such a thing exists!)

Re: Sakana Fugu

#106

Post by blks »

bicx wrote: post #104
holistio wrote: post #23

You pay $200/month to Anthropic, $200/month to OpenAI, $200/month to Cursor, $200/month to $200/month to Google, and seeing that it didn't come to a nice round $1024/month, you pay $200/month to Sakana to coordinate it all, because why not. While you're at it, feel free to send me $200 as well, I'll generate a crypto address ending with "AI".

I wish I only paid $200/mo for Anthropic! Multiply that by 20x.

What are you getting out of it at $4000/month?

Re: Sakana Fugu

#107

Post by ricardobeat »

holistio wrote: post #23

You pay $200/month to Anthropic, $200/month to OpenAI, $200/month to Cursor, $200/month to $200/month to Google, and seeing that it didn't come to a nice round $1024/month, you pay $200/month to Sakana to coordinate it all, because why not. While you're at it, feel free to send me $200 as well, I'll generate a crypto address ending with "AI".

My current setup:

    $20/month: Claude Code
    $10/month: Minimax
    $16/month: Xiaomi Mimo
    $10/month: Opencode Go
Opus at low/medium effort generates plans. Then several coordinator/worker pairs are possible: DeepSeek v4 Pro + Minimax M3, Mimo v2.5 Pro + Mimo v2.5, Mimo + Minimax, Sonnet 4.6 + Haiku. I've been running hundreds of long multi-agent sessions, topped up extra credits here and theere, but haven't reached $200/month spend yet. Relying entirely on Claude/Codex feels like a waste of cash now.

Re: Sakana Fugu

#108

Post by andai »

ljlolel wrote: post #96
mark_l_watson wrote: post #92

Nice idea but expensive. It looks like they don’t add very low cost models like DeepSeek v4 flash into their mix. After a few months of spending money on the best frontier models, now I am spending time using DeepSeek v4 flash as my workhorse, and flipping to more capable (but still very inexpensive) open models on an as-needed basis. We all make our own tool selection decisions, but for me, I feel happier and enjoy…

We found that an all open source fusion was 1/3 the price and better than Fable https://trustedrouter.com/blog/open-fusion-beats-fable-5

Brilliant. What this actually is, is a swarm, albeit a very small one. I'm wondering if for research specifically, swarm size (on higher temp?) would outweigh model size.

At least, for the initial data gathering phase. You'd probably want a sequence of progressively larger models to filter it.

Have you guys tested it on anything other than research?

Re: Sakana Fugu

#109

Post by qainsights »

rvz wrote: post #25
holistio wrote: post #23

You pay $200/month to Anthropic, $200/month to OpenAI, $200/month to Cursor, $200/month to $200/month to Google, and seeing that it didn't come to a nice round $1024/month, you pay $200/month to Sakana to coordinate it all, because why not. While you're at it, feel free to send me $200 as well, I'll generate a crypto address ending with "AI".

Pay $0 to run a local model or even a cheap DeepSeek V4 model via their API which is close to free per million tokens. These prices are just going to get raced to $0.

Not everyone can run local models. It is also expensive will be outdated soon as the model evolves.

Re: Sakana Fugu

#110

Post by andai »

NitpickLawyer wrote: post #48
GolfPopper wrote: post #15

This is a joke, right?

Not necessarily. There were some tests last year-ish from hf that showed that simply alternating (randomly) between claude and gpt (whatever their versions were at the time) on a task produced better results than either of them individually. So during a task, the first call was sent to one, then the other and so on. There's also the concept of "smart routing" requests based on some heuristics / embeddings. You'd get…

See also: Agents built from alloys (July 2025)

https://news.ycombinator.com/item?id=44630724

They randomly alternated between frontier LLMs and got a massive boost to performance on cybersecurity tasks.

Post reply on HN