Hacker News: byefruit

New comment by byefruit in "Building more with GPT-5.1-Codex-Max"

byefruit — Wed, 19 Nov 2025 19:33:23 +0000

I've just found myself using OpenRouter if we need Google models for a project, it's worth the extra 5% just not to have to deal with the utter disaster that is their product offering.

New comment by byefruit in "Why I love OCaml (2023)"

byefruit — Fri, 07 Nov 2025 17:09:26 +0000

This is the wrong interpretation of the oxcaml project. If you look at the features and work on it, it's primarily performance or parallelism safety features.

The latter going much further than most mainstream languages.

New comment by byefruit in "Australia has so much solar that it's offering everyone free electricity"

byefruit — Thu, 06 Nov 2025 17:15:35 +0000

7.3% return, not bad. As battery prices drop it will get even better.

New comment by byefruit in "Gemini 3.0 Pro – early tests"

byefruit — Thu, 02 Oct 2025 19:13:07 +0000

And even when it does copy other products, it seems to be doing a terrible job of them.

Google's AI offering is a complete nightmare to use. Three different APIs, at least two different subscriptions, documentation that uses them interchangeably.

For Gemini's API it's often much simpler to actually pay OpenRouter the 5% surchargeto BYOK than deal with it all.

I still can't use my Google AI Pro account with gemini-cli..

New comment by byefruit in "Jules, our asynchronous coding agent"

byefruit — Wed, 06 Aug 2025 17:56:00 +0000

How is this different from https://github.com/google-gemini/gemini-cli ?

Edit: it seems this is a hosted version. Would be nice if they actually joined up some of their products.

New comment by byefruit in "Qwen3-4B-Thinking-2507"

byefruit — Wed, 06 Aug 2025 17:30:22 +0000

The openrouter rankings can be biased.

For example, Google's inexplicable design decisions around libraries and APIs means it's often worth the 5% premium to just use OpenRouter to access their models. In other cases it's about which models particular agents default to.

Sonnet 4 is extremely good for tool-usage agentic setups though - something I have found other models struggle to do over a long-context.

New comment by byefruit in "The U.K. closed a tax loophole for the global rich, now they're fleeing"

byefruit — Sat, 19 Jul 2025 22:49:26 +0000

Before just accepting this at face value, New Statesman claim this is not the case:

https://www.newstatesman.com/politics/2025/07/the-british-we...

New comment by byefruit in "Mistral ships Le Chat – enterprise AI assistant that can run on prem"

byefruit — Wed, 07 May 2025 19:01:39 +0000

You are probably getting downvoted because you don't give any model generations or versions ('ChatGPT') which makes this not very credible.

New comment by byefruit in "Google Gemini has the worst LLM API"

byefruit — Sun, 04 May 2025 18:28:22 +0000

100% this. We actually use OpenRouter (and pay their surcharge) with Gemini 2.5 Pro just because we can actually control spend via spent limit on keys (A++ feature) and prepaid credit.

New comment by byefruit in "Fossil fuels fall below 50% of US electricity for the first month on record"

byefruit — Mon, 21 Apr 2025 12:31:42 +0000

Indeed, average in CA is $260/month so $5k pays off very fast in some places.

New comment by byefruit in "Gemini 2.5 Flash"

byefruit — Thu, 17 Apr 2025 19:27:12 +0000

It's interesting that there's a price nearly 6x price difference between reasoning and no reasoning.

This implies it's not a hybrid model that can just skip reasoning steps if requested.

Anyone know what else they might be doing?

Reasoning means contexts will be longer (for thinking tokens) and there's an increase in cost to inference with a longer context but it's not going to be 6x.

Or is it just market pricing?

New comment by byefruit in "AWS announces 85% price reductions for S3 Express One Zone"

byefruit — Mon, 14 Apr 2025 16:00:43 +0000

"In addition, S3 Express One Zone has reduced the per-GB charges for data uploads and retrievals by 60 percent, and these charges now apply to all bytes transferred rather than just portions of requests greater than 512 KB"

It's not clear but are there cases where this could be a significant price rise? If you exclusively had small objects (<512kb) being written and read then this could add up quickly.

New comment by byefruit in "Skywork-OR1: new SOTA 32B thinking model with open weight"

byefruit — Sun, 13 Apr 2025 16:38:49 +0000

> Both of our models are trained on top of DeepSeek-R1-Distill-Qwen-7B and DeepSeek-R1-Distill-Qwen-32B.

Not to take away from their work but this shouldn't be buried at the bottom of the page - there's a gulf between completely new models and fine-tuning.

New comment by byefruit in "It's five grand a day to miss our S3 exit"

byefruit — Sun, 30 Mar 2025 12:53:11 +0000

https://aws.amazon.com/snowball/pricing/ snowball seems to support getting data out of S3 though you still end up paying extortionate egress charges.

New comment by byefruit in "Meta keeps 'block' lists of ex-employees – and even a VP can't get you off them"

byefruit — Thu, 06 Mar 2025 21:26:55 +0000

I think this is where (in the EU) a Subject Access Request could work.

New comment by byefruit in "How the U.K. broke its own economy"

byefruit — Tue, 04 Mar 2025 08:32:01 +0000

This has already been proposed by the current government for wind farms: https://inews.co.uk/news/environment/new-energy-bill-discoun...

And the switch to zonal energy pricing will likely have a similar effect for other sources of generation.

New comment by byefruit in "Long-Context GRPO"

byefruit — Fri, 21 Feb 2025 10:35:37 +0000

I'm waiting for https://github.com/huggingface/trl/pull/2810 to land. I think this should work with the existing unsloth setup without changes.

New comment by byefruit in "DeepRAG: Thinking to retrieval step by step for large language models"

byefruit — Tue, 04 Feb 2025 19:07:25 +0000

> This generally requires thousands of examples created by an expert in the field.

Or an AI model pretending to be an expert in the field... (works well in a few niche domains I have used this in)

New comment by byefruit in "O3-mini System Card [pdf]"

byefruit — Fri, 31 Jan 2025 18:27:28 +0000

I look forward to people applying the same standards to the OpenAI's O3 as they did Deepseek's R1 release and paper in the discussions last week.

New comment by byefruit in "DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL"

byefruit — Sat, 25 Jan 2025 20:32:20 +0000

That's not what I'm saying, they may be hiding their true compute.

I'm pointing out that nearly every thread covering Deepseek R1 so far has been like this. Compare to the O1 system card thread: https://news.ycombinator.com/item?id=42330666

Very different standards.