Bob

[Bob @Bob88488488]

@0x_Northland 告诉你个秘密，大部分中转站都是luna，被降智了

[Bob @Bob88488488]

@NFT_Chen 兄弟，DeepSWE v1.1 V4.1 Flash是74%，转挑跑分打得过的比，无敌了，拉完了

[Bob @Bob88488488]

@thsottiaux Fix our usage limit!

[Bob @Bob88488488]

@Johanne61545102 @AdamHoltererer The usage limit is the only problem. Even with 6-Sol, Codex Plus has a much lower usage limit, about 1/3 of Claude Pro when using Opus 5.5. That’s pathetic!

[Bob @Bob88488488]

@MingTian685474 @QIJIUnot79 降价了但是额度跟着一起降

[Bob @Bob88488488]

@thsottiaux Your $100 plan has limits similar to Claude’s $20 plan. Fix it!

[Bob @Bob88488488]

@Curline1222 Already switched. Downgraded my Pro to Plus and moved to Claude Pro. $20 feels like $100 to me. I keep Plus because GPT is better at some hard STEM questions than Claude, but otherwise Claude just mogged GPT, even Sonnet 5.5 is better than GPT-6-Sol

[Bob @Bob88488488]

@BohuTANG 是5.6-luna不是5.5

[DeepSeek @deepseek_ai]

🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.

🔹 Introducing the smallest model in our new architecture family, with native visual understanding.
🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models.

1/6

[adi @adonis_singh]

this is astra btw.

first model to visibly leave me at a loss for words

[Kimi.ai @Kimi_Moonshot]

Introducing Kimi K3: Open Frontier Intelligence

🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal
🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts
🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost
🔹 Built for long-horizon agentic coding and self-evolving workflows

Kimi K3 is now live on on https://t.co/zrk6zZxZUo, Kimi Work, Kimi Code, and the Kimi API.
Open Weights by July 27, 2026.

🔗 API: https://t.co/XCrgjXAqMw
🔗 Tech blog: https://t.co/YTfiMSNM1f

[Da7em @Da7_Tech]

Same task. Same prompt. Same setup.

Opus 4.8 completed the task.
It used 2% of the weekly limit and 7% of the 7-hour limit.
GLM 5.2 did not complete the task.
It consumed 20% of the weekly limit and 100% of the 5-hour limit.

Token usage is completely abnormal:
Opus 4.8: less than 1.5M tokens

GLM 5.2: reached 53M tokens

And this is despite the app showing only around 1.67M tokens.
This means one thing very clearly: cached tokens are not working properly on GLM 5.2. The repeated context is being counted as normal input instead of cached tokens.

I have been dealing with this issue for more than a month with no real solution.

By the way, I am on the Pro plan, and I paid for a full yearly subscription.

These numbers are not normal. This is not usage behavior. This is a serious billing/cache accounting problem on GLM 5.2.

[Bob @Bob88488488]

@anfflee @plantegg GPT-3 175B
2020年5月就发布了
盘古2020年3月 才立项

[Bob @Bob88488488]

@Balder13946731 大模型订阅更像卖保险

[yv @yvbbrjdr]

推荐大家读一下MAI-Thinking-1的technical paper，里面有详细的怎么训出一个SOTA LLM的（几乎）所有细节。

https://t.co/it5mCFd6v3

[Bob @Bob88488488]

@Balder13946731 其实就是 老登 太多

[Bob @Bob88488488]

@Balder13946731 恐怖的是即使涨40%这样PE也只有24

[Bob @Bob88488488]

@realNyarime A社太坏了，中文蒸馏Qwen，让中文用户永远只能用Qwen

[Bob @Bob88488488]

@realNyarime @lei_ren38662 Gemini是美国豆包，这蒸馏完不真成 哈基米 了…