Cipher Vectors @CipherVectors

Agentic Coding Dubai Joined September 2012

Tweets

2K
Followers

381
Following

1K
Likes

3K

Cipher Vectors @CipherVectors

2 days ago

The problem with GLM 4.5 is the 128k context window. You cant even implement a 40-task feature ! it keeps compacting and loosing the track. they must work on it. Another issue is that it gives up in the middle of tasks without throwing any error or reason!

0 0 0 37 0

Cipher Vectors @CipherVectors

a week ago

Claude Sonnet 4 ( Thinking or non thinking Model doesnt matter) works the best with Playwright MCP: follows the instructions precisely. Try it

1 0 0 24 0

Cipher Vectors @CipherVectors

2 weeks ago

F..k Windsurf: it is no longer usable ! it is just wasting my credits by saying "This may be stuck. Please Try cancelling...." . Windsurf has stolen too much from the customers!

0 1 1 72 0

Cipher Vectors @CipherVectors

2 weeks ago

until grok code fast 1 is free, no GLM 4.5 or QWEN !

0 0 0 109 0

Cipher Vectors @CipherVectors

3 weeks ago

After LLM build a feature, then you start asking questions. at this point the LLM treat your questions as requiements because it knows that these are all the flaws available in the pipeline.

0 0 0 21 0

Cipher Vectors @CipherVectors

3 weeks ago

What makes your business model resilient against competition from AI tech giants who have more resources and faster scaling capabilities?

0 0 0 25 0

Cipher Vectors @CipherVectors

3 weeks ago

Sonnet 4 is the gladiator

0 0 0 15 0

Cipher Vectors @CipherVectors

3 weeks ago

GLM 4.5 is faster than Qwen-3-Max Preview Test and compare them here chat.qwen.ai and chat.z.ai

0 0 0 73 0

Cipher Vectors @CipherVectors

3 weeks ago

if your coding ai agent is stuck in a loop hole, keep asking what is it trying to do . then you can direct it what to do: just like supervising junior dev . Happy Coding

0 0 0 29 0

Cipher Vectors @CipherVectors

3 weeks ago

right now everyone on X is shilling the GLM-4.5 but no one is there to shwo a video or images of the prompt and the results of GLM 4.5 vs KIMI K2 vs QWEN 3 vs Sonnet 4 vs GPT-5 series