[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

@KLieret Avatar @KLieret Kilian Lieret

Kilian Lieret posts on X about leaderboard, more info, at least, if you the most. They currently have XXXXX followers and XX posts still getting attention that total XXXXXX engagements in the last XX hours.

Engagements: XXXXXX #

Engagements Line Chart

Mentions: X #

Mentions Line Chart

Followers: XXXXX #

Followers Line Chart

CreatorRank: XXXXXXX #

CreatorRank Line Chart

Social Influence

Social category influence exchanges XXXX%

Social topic influence leaderboard #12, more info 9.09%, at least 9.09%, if you 9.09%, the official 9.09%, claude 9.09%, funny 9.09%, trade 9.09%, gemini XXXX%

Top accounts mentioned or mentioned by @carlosejimenez @jyangballin @ofirpress @dronathon @hiimg2html

Top Social Posts

Top posts by engagements in the last XX hours

"Updates to official SWE-bench leaderboard: GPT XXX high is #3 after Gemini for similar price point. GPT XXX medium reduces gap to Sonnet XXX and is much cheaper. All independently evaluated using identical mini-swe-agent setup. More info & full trajectories in 🧵"
X Link 2025-12-11T21:00Z 1313 followers, 72K engagements

"GPT XXX get near max performance for step limits as small as 50-ish (much less than Gemini or Opus that require XXX steps or more to get max performance)"
X Link 2025-12-11T21:00Z 1313 followers, 2786 engagements

"For minimax in particular you should use a very high step limit to get the full performance (at least 200). For deepseek limits below XXX steps also seem viable"
X Link 2025-12-02T20:14Z 1192 followers, XXX engagements

"You can also use step limits to trade cost vs performance. If you don't care about long latency deepseek is hard to beat otherwise GLM is very cost efficient"
X Link 2025-12-02T20:14Z 1192 followers, XXX engagements

"Updates to the official SWE-bench leaderboard: Minimax M2 is best open source model (but expensive). Deepseek v3.2 reasoning close behind very cheap but very slow. GLM XXX reaches good performance fast and cheap. All independently evaluated with mini-swe-agent. Details in 🧵"
X Link 2025-12-02T20:14Z 1313 followers, 117K engagements

"GPT models continue to use significantly less steps (impressively just XX for medium/17 for high) than Gemini and Claude models"
X Link 2025-12-11T21:00Z 1313 followers, 3835 engagements

"But also frequently some funny first actions frequently using the Gregorian alphabet version of ls -la"
X Link 2025-12-11T21:00Z 1313 followers, 2950 engagements

"Varying step limits we can trade performance vs cost. GPT is one of the most cost-efficient models and this trend continues with GPT XXX while to get near-maximum performance Gemini is still the cheapest"
X Link 2025-12-11T21:00Z 1313 followers, 2423 engagements