🏆 Model Leaderboard

Compare performance across dimensions to quickly find the best open model for you

UPDATE
🥇
Claude Opus 5 Max New #1
1000K context Proprietary
Overall score 100.0 ★★★★★
🥈
Kimi K3 Max
1049K context Kimi K3 license
Overall score 98.2 ★★★★★
🥉
Claude Opus 5
1000K context Proprietary
Overall score 97.0 ★★★★★
4
Claude Fable 5
1000K context Proprietary
Overall score 92.9 ★★★★★
5
GPT-5.6 Sol
1050K context Proprietary
Overall score 92.3 ★★★★★
6
GLM-5.2 Max
1000K context MIT
Overall score 88.7 ★★★★☆
7
Claude Opus 4.8
1000K context Proprietary
Overall score 86.8 ★★★★☆
8
Claude Opus 4.7
1000K context Proprietary
Overall score 86.0 ★★★★☆
9
Grok 4.5
500K context Proprietary
Overall score 85.0 ★★★★☆
10
Claude Opus 4.6
1000K context Proprietary
Overall score 84.6 ★★★★☆
11
Claude Sonnet 5
1000K context Proprietary
Overall score 84.5 ★★★★☆
12
Muse Spark 1.1
Proprietary
Overall score 83.7 ★★★★☆
13
Seed 2.1 Pro
Proprietary
Overall score 82.8 ★★★★☆
14
Gemini 3.6 Flash
1049K context Proprietary
Overall score 82.7 ★★★★☆
15
Claude Sonnet 4.6
1000K context Proprietary
Overall score 82.3 ★★★★☆
16
GLM-5.1
203K context MIT
Overall score 81.8 ★★★★☆
17
Hy3
262K context Apache 2.0
Overall score 81.8 ★★★★☆
18
Qwen 3.7 Max
1000K context Proprietary
Overall score 81.7 ★★★★☆
19
Kimi K2.6
262K context Modified MIT
Overall score 81.0 ★★★★☆
20
GPT-5.5
Proprietary
Overall score 80.7 ★★★★☆
21
Claude Opus 4.5
200K context Proprietary
Overall score 79.4 ★★★★☆
22
Minimax M3
MiniMax Community License
Overall score 79.3 ★★★★☆
23
Gemini 3.5 Flash
1049K context Proprietary
Overall score 79.3 ★★★★☆
24
Qwen3.6 Max
262K context Proprietary
Overall score 77.8 ★★★★☆
25
MiMo 2.5 Pro
1050K context MIT
Overall score 77.4 ★★★★☆
26
Kimi K2.7 Code
262K context Modified MIT
Overall score 77.3 ★★★★☆
27
Deepseek V4 Pro
1049K context MIT
Overall score 76.4 ★★★★☆
28
GPT-5.4
1050K context Proprietary
Overall score 76.2 ★★★★☆
29
Qwen3.6 Plus
1000K context Proprietary
Overall score 75.8 ★★★★☆
30
Gemini 3.5 Flash Lite
1049K context Proprietary
Overall score 75.3 ★★★★☆
31
Gemini 3.1 Pro
1049K context Proprietary
Overall score 74.6 ★★★★☆
32
Gemini 3 Pro
1049K context Proprietary
Overall score 73.8 ★★★★☆
33
Gemini 3 Flash
1049K context Proprietary
Overall score 73.8 ★★★★☆
34
Mimo V2.5
1050K context MIT
Overall score 73.6 ★★★★☆
35
Kimi K2.5
Modified MIT
Overall score 73.6 ★★★★☆
36
GLM-5
203K context MIT
Overall score 73.5 ★★★★☆
37
GLM-4.7
205K context MIT
Overall score 73.3 ★★★★☆
38
Mimo V2 Pro
1049K context Proprietary
Overall score 73.3 ★★★★☆
39
GPT-5.2
400K context Proprietary
Overall score 71.9 ★★★★☆
40
GPT-5
400K context Proprietary
Overall score 71.9 ★★★★☆
41
Inkling
1049K context Apache 2.0
Overall score 71.8 ★★★★☆
42
GPT-5.3 Codex
400K context Proprietary
Overall score 70.9 ★★★★☆
43
GLM-5v
203K context Proprietary
Overall score 70.1 ★★★★☆
44
Qwen3.5 397b A17b
262K context Apache 2.0
Overall score 70.1 ★★★★☆
45
GPT-5.4 Mini
400K context Proprietary
Overall score 69.9 ★★★☆☆
46
Minimax M2.7
205K context Modified MIT
Overall score 69.7 ★★★☆☆
47
Claude Sonnet 4.5
200K context Proprietary
Overall score 69.2 ★★★☆☆
48
GPT-5.1
400K context Proprietary
Overall score 69.0 ★★★☆☆
49
Claude Opus 4.1
200K context Proprietary
Overall score 68.9 ★★★☆☆
50
Minimax M2.1
205K context MIT
Overall score 68.8 ★★★☆☆

CURRENT MODEL

Claude Opus 5 Max

Rank #1

Ranked #1 in the LMSYS Arena Coding evaluation, overall score 100.0.

Arena score: 100.0 (raw 1725)
Context window: 1M Tokens
Organization: Anthropic
License: Proprietary
Official price Input $5.00 / 1M tokens Output $25.00 / 1M tokens

SELECT ACTIVE GATEWAYS

● Encrypted link verified
All data comes from public testing; actual results may vary by usage and scenario.