Skip to content
LLM Pulse

Leaderboard

SciCode

30 models scored · metric: percent correct. 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.Step 3.5 Flash284
  2. 2.Step 3.5 Flash 2603257
  3. 3.Gemini 2.5 Flash-Lite193
  4. 4.GPT-4o mini174
  5. 5.Mistral Small 4145
#ModelScoreBest in /MBest out /MPts / $Context
🥇Fugusakana
60.1
1M
🥈Fugu Ultrasakana
58.7
$5$305.21M
🥉Qwen3.7 Maxalibaba
53.5
$2.50$7.5014.31M
4Gemini 2.5 Progoogle
42.8
$0.87$717.81.05M
5GPT-5-Codexopenai
40.9
$1.10$913.3400K
6Step 3.5 Flashstepfun
40.4
$0.09$0.3284256K
7Step 3.7 Flashstepfun
40
$0.185$1.1196.1256K
8Gemini 2.5 Flashgoogle
39.4
$0.21$1.8064.91.05M
9Step 3.5 Flash 2603stepfun
38.5
$0.1$0.3257256K
10GLM-4.6zhipuai
38.4
$0.6$2.2038.4205K
11Qwen3 Maxalibaba
38.3
$1.20$616.0262K
12Mistral Small 4mistral
38
$0.15$0.6145256K
13Mistral Large 3mistral
36.2
$0.5$1.5048.3262K
14DeepSeek-R1deepseek
35.7
$0.7$2.5031.0128K
15GLM-4.5zhipuai
34.8
$0.6$2.2034.8131K
16GPT-4o (2024-11-20)openai
33.3
$2.50$107.6128K
17Devstral 2mistral
33.1
$0.4$241.4262K
18Mistral Medium 3mistral
33.1
$0.4$241.4131K
19GPT-4o (2024-08-06)openai
33.1
$2.50$107.6128K
20GPT-4 Turboopenai
31.9
$9$272.4128K
21GPT-4o (2024-05-13)openai
30.9
$5$154.1128K
22GLM-4.5-Airzhipuai
30.6
$0.2$1.1072.0131K
23Mistral Large 2.1mistral
29.2
$2$69.7131K
24Qwen3-Coder 30B-A3B Instructalibaba
27.8
$0.45$2.2530.9262K
25Llama-3.3-70B-Instructmeta
26
$0.22$0.589.7128K
26GPT-4o miniopenai
22.9
$0.075$0.3174128K
27Sonarperplexity
22.9
$0.25$2.5028.2128K
28Sonar Properplexity
22.6
$3$153.8200K
29GLM-4.5Vzhipuai
22.1
$0.6$1.8024.664K
30Gemini 2.5 Flash-Litegoogle
19.3
$0.1$0.11931.05M

Cheapest scorer: GPT-4o mini at $0.075 input /M.