Matching posts

Matching terms are highlighted

Grok Bot may route models by exit IP, test suggests

…的是，我用新加坡 IP 和美国家宽 IP 测了下，前者 grok-4.5 置信度 95%，后者 opus-5.5 置信度 99% 样本不具备参考性，想看看自己的 Grok Bot 背后路由了什么模型，可以测试…

grok@wangray　Filed in No.012

1 day agoCopy original

POST

GPT-6 Astra trails Opus 5.5, and Codex is losing this user

…줄 알았나보다. 완전 착각이었다. 그 어느것하나 앞서는게 없다. GPT-6-Astra<Opus-5.5 Dots<Muse<<Grok Bot Space는 진짜 모르겠다. 내가 클로드 코드에서…

gpt-6@minheo93698092　Filed in No.001

1 day agoCopy original

POST

Apple benchmark: Sonnet-5.5 tops, Gemini 4 Argon sixth

… three boards on Apple's new benchmark ahead of Opus-5.5 and Fable-5.1, while Gemini 4 Argon placed sixt…

gemini-4@loveAIwang　Filed in No.008

2 days agoCopy original

Opus 5.5 Max renders a video from pure code on the same prompt

Claude Opus-5.5 Max, same prompt. Pure code.

claude@diegocabezas01　Filed in No.010

2 days agoCopy original

Cognition's FrontierCode 2.0: DeepSeek V4.1 Flash 19.3% vs Opus 5.5 45.5%

The new coding benchmark from 20+ maintainers leaves requirements implicit and zeroes hard-rule violations; DeepSeek V4.1 Flash scores 19.3% against Opus 5.5's 45.5%.

deepseek-v4.1@Moblacksoil　Filed in No.020

15 hours agoCopy original

POST

DeepSeek V4.1 Flash in Cursor: CursorBench 4.0 scores and per-task cost by thinking level

On CursorBench 4.0, DeepSeek V4.1 Flash scores 44.9% at $0.60 per task on Low up to 47.7% at $0.98 on Max, while Opus 5.5 Low gets 43.7% at $1.17.

deepseek-v4.1@nextaurame　Filed in No.020

11 hours agoCopy original

Intelligence Index 5.0: DeepSeek V4.1 Flash scores 32, 26 behind Opus 5.5

The ranking post puts Opus 5.5 at 58, GPT-6 Astra at 57 and GPT-6.1 Sol at 55, with DeepSeek V4.1 Flash and GLM-5.3 tied at 32, and adds a coding test judged by real repo maintainers.

deepseek-v4.1@MartinM91217805　Filed in No.020

9 hours agoCopy original

POST

Commenter calls DeepSeek V4.1 Flash 1017 a hypothetical Opus 5.3

A commenter finds the comparison of DeepSeek V4.1 Flash 1017 to a nonexistent Opus 5.3 wild, while acknowledging the current hype around open-weight models.

deepseek-v4.1@Blum_OG　Filed in No.020

16 hours agoCopy original

POST

Matthew Berman likens DeepSeek V4.1 Flash 1017 to a hypothetical Opus 5.3

A subjective comparison placing DeepSeek V4.1 Flash 1017 at roughly the level of a hypothetical Opus 5.3, without benchmark data.

deepseek-v4.1@MatthewBerman　Filed in No.020

16 hours agoCopy original

POST

After Astra nerf, user returns to Opus 5.5 and latest DeepSeek V4.1 Flash

The user says Astra got nerfed, so they went back to Opus 5.5, Fable 5.1 and the latest DeepSeek V4.1 Flash.

deepseek-v4.1@smakosh　Filed in No.020

16 hours agoCopy original

Showing10of60posts

10 per batch

Keep scrolling to reveal posts 11–20

Showing the first 60 matches. Narrow the query with a model name or tag to find more specific results.