Hacker News: mekpro

New comment by mekpro in "Claude Fable 5 by Anthropic, releasing tomorrow"

mekpro — Tue, 09 Jun 2026 04:48:48 +0000

source ?

New comment by mekpro in "Meta Keeps Delaying the Release of Its New AI Model to Developers"

mekpro — Sat, 06 Jun 2026 14:56:57 +0000

API server is not hard problem and not make sense for indefinite postpone. I think the more likely explanation is model quality.

Too bad for Meta, and very sad day Llama.

Meta Keeps Delaying the Release of Its New AI Model to Developers

mekpro — Sat, 06 Jun 2026 12:15:59 +0000

Article URL: https://www.wsj.com/tech/ai/meta-keeps-delaying-the-release-of-its-new-ai-model-to-developers-f8569c8c

Comments URL: https://news.ycombinator.com/item?id=48424252

Points: 67

# Comments: 26

New comment by mekpro in "MAI-Code-1-Flash"

mekpro — Wed, 03 Jun 2026 14:44:38 +0000

The technical report is very detailed and would 'reinforcement learning' of future researchers, Thanks Microsoft!

New comment by mekpro in "Expanding Project Glasswing"

mekpro — Tue, 02 Jun 2026 15:30:33 +0000

Yes, 300 MW from SpaceX helps a lot, but I think that’s mainly to support Opus demand, which has grown faster than expected. If Mythos is roughly 5× more expensive to serve than Opus, as the pricing suggests, then 300 MW is nowhere near enough to enable large-scale deployment of Mythos.

As an ordinary developer who relies on a $20–$200/month subscription, I feel disappointed by the release of a paper describing a model that I can’t actually use.

New comment by mekpro in "Expanding Project Glasswing"

mekpro — Tue, 02 Jun 2026 14:52:28 +0000

It’s clear that Anthropic has run out of the compute capacity needed to serve Mythos publicly.

They’re using security concerns to mask their inability to deliver the model at scale, while still trying to maintain their lead over OpenAI. As a result, they’ve chosen to release it privately under the banner of an “ethical” rollout.

New comment by mekpro in "Kimi Claw"

mekpro — Mon, 16 Feb 2026 03:06:20 +0000

Opus is definitely in its own league. I use Kimi/Gemini-cli code regularly to save cost and from my experience, Kimi 2.5 is more solid than Gemini Flash 3.0 for coding. While Gemini Flash 3.0 is generally faster, it usually break the syntax and skip important prompt. Kimi 2.5 can write very good code and can plan very well.

New comment by mekpro in "Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model"

mekpro — Tue, 27 Jan 2026 14:59:52 +0000

Except that, In OpenRouter, Deepseek always maintain in Top 10 Ranking. Although I did not use it personally, i believe that their main advantage over other model is price/performance.

New comment by mekpro in "Apple is fighting for TSMC capacity as Nvidia takes center stage"

mekpro — Fri, 16 Jan 2026 03:32:28 +0000

I think the opposite. Having NVIDIA investing in TSMC's bleeding-edge process node should benefit Apple rather than disadvantage.

It means that Apple doesn't have to be sole investor in latest node development which is more harder to justify, especially in the year where smartphone upgrade cycle is slowdown. Having NVIDIA (and AI boom) in the picture should help Apple reduce CAPEX for their semi-conductor investment.

New comment by mekpro in "1300 Still Images from the Animated Films of Hayao Miyazaki's Studio Ghibli (2023)"

mekpro — Sat, 13 Dec 2025 05:46:11 +0000

They are so beautiful that i dont want any of these been stole by AI.

New comment by mekpro in "DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning [pdf]"

mekpro — Fri, 28 Nov 2025 01:39:24 +0000

How this improvement translate into real world agentic coding task ?

New comment by mekpro in "Open models by OpenAI"

mekpro — Wed, 06 Aug 2025 00:30:56 +0000

i got 70 token/s on m4 max

New comment by mekpro in "Open models by OpenAI"

mekpro — Wed, 06 Aug 2025 00:29:33 +0000

try enable flash attention and offload all layer to GPU

New comment by mekpro in "Claude Code weekly rate limits"

mekpro — Tue, 29 Jul 2025 09:08:49 +0000

Is this limit will also count together with Claude Chat ?

New comment by mekpro in "OpenAI’s Windsurf deal is off, and Windsurf’s CEO is going to Google"

mekpro — Sat, 12 Jul 2025 06:42:58 +0000

you can easily reach 50$ per day. by force switching model to opus /model opus it will continue to use opus eventhough there is a warning about approaching limit.

i found opus is significantly more capable in coding than sonnet, especcially for the task that is poorly defined, thinking mode can fulfill alot of missing detail and you just need to edit a little before let it code.

New comment by mekpro in "Gemini CLI"

mekpro — Wed, 25 Jun 2025 14:28:38 +0000

Just refactored 1000 lines of Claude Code generated to 500 lines with Gemini Pro 2.5 ! Very impressed by the overall agentic experience and model performance.

New comment by mekpro in "Ask HN: How to learn CUDA to professional level"

mekpro — Sun, 08 Jun 2025 15:32:42 +0000

To professionals in the field, I have a question: what jobs, positions, and companies are in need of CUDA engineers? My current understanding is that while many companies use CUDA's by-products (like PyTorch), direct CUDA development seems less prevalent. I'm therefore seeking to identify more companies and roles that heavily rely on CUDA.

New comment by mekpro in "Devstral"

mekpro — Thu, 22 May 2025 14:11:56 +0000

it can use tool to explore directory like ls grep out of the box.

Judge said Meta illegally used books to build its AI

mekpro — Mon, 05 May 2025 11:16:16 +0000

Article URL: https://www.wired.com/story/meta-lawsuit-copyright-hearing-artificial-intelligence/

Comments URL: https://news.ycombinator.com/item?id=43893762

Points: 406

# Comments: 341

New comment by mekpro in "Gemma 3 QAT Models: Bringing AI to Consumer GPUs"

mekpro — Sun, 20 Apr 2025 15:45:59 +0000

Gemma 3 is way way better than Llama 4. I think Meta will start to lose its position in LLM mindshare. Another weakness of Llama 4 is its model size that is too large (even though it can run fast with MoE), which greatly limits the applicable users to a small percentage of enthusiasts who have enough GPU VRAM. Meanwhile, Gemma 3 is widely usable across all hardware sizes.