Skip to content
LLM Pulse

Leaderboard

SWE-Atlas Codebase QnA

26 models scored · metric: score. Prices are the cheapest listed across all serving providers. Pts/$ = score ÷ blended price (3×input + output) / 4 — how much score each dollar buys.

Best value · top 5 by pts per dollar
  1. 1.DeepSeek V4 Pro125
  2. 2.Kimi K2.658.3
  3. 3.GLM-5.134.0
  4. 4.MiniMax-M2.519.6
  5. 5.Kimi K2.518.7
#ModelScoreBest in /MBest out /MPts / $Context
🥇Claude Opus 4.7anthropic
81
$5$258.11M
🥈GPT-5.5openai
80.8
$4.55$27.277.91.05M
🥉GPT-5.5openai
79.1
$4.55$27.277.71.05M
4Claude Opus 4.7anthropic
78.4
$5$257.81M
5GPT-5.5openai
75
$4.55$27.277.31.05M
6GLM-5.1zhipuai
73.2
$1.40$4.4034.0200K
7GPT-5.4openai
72.9
$2.20$1414.21.05M
8GPT-5.4openai
72.4
$2.20$1414.11.05M
9Claude Opus 4.6anthropic
71.9
$5$257.21M
10Claude Opus 4.7anthropic
71.7
$5$257.21M
11Claude Sonnet 4.6anthropic
70.3
$3$1511.71M
12DeepSeek V4 Prodeepseek
67.8
$0.435$0.871251M
13Kimi K2.6moonshotai
59.8
$0.5$2.6058.3262K
14Gemini 3.1 Pro Previewgoogle
45.6
$2$1210.11.05M
15GPT-5.5openai
45.43
$4.55$27.274.41.05M
16GPT-5.4openai
40.8
$2.20$147.91.05M
17GPT-5.4openai
36.3
$2.20$147.01.05M
18Claude Opus 4.6anthropic
33.3
$5$253.31M
19GPT-5.3 Codexopenai
32.6
$1.60$137.3400K
20Claude Sonnet 4.6anthropic
31.2
$3$155.21M
21Claude Opus 4.6anthropic
30
$5$253.01M
22GLM-5zhipuai
20.5
$1$3.2013.2205K
23Gemini 3.1 Pro Previewgoogle
13.5
$2$123.01.05M
24Kimi K2.5moonshotai
13.1
$0.3$1.9018.7262K
25MiniMax-M2.5minimax
10.3
$0.3$1.2019.6205K
26Gemini 3 Flash Previewgoogle
8.2
$0.5$37.31.05M

Cheapest scorer: Kimi K2.5 at $0.3 input /M.