Crimson AI NewsA CrimsonLingua Network service
EN ع
Benchmarks

AI Model Benchmarks

Independent model performance, based on Artificial Analysis.

# Model Intelligence Speed (t/s) Latency (s) Price ($/1M) Context
1
Claude Opus 5.5 (max with fallback)
AAnthropic
57.6
— — — 1M
2
Claude Fable 5.1 (max with fallback)
AAnthropic
53.4
69 — — 1M
3
GPT-6 Astra (max)
OOpenAI
52.7
54 — — 1M
4
Muse Spark 1.3 (max)
48.1
222 36.07 $2.00 1M
5
GPT-6 Sol (max)
OOpenAI
47.5
99 181.31 $4.00 872K
6
Grok 4.7 (xhigh)
XxAI
46.4
40 0.95 $3.00 5K
7
MiMo-V2.6-Pro
46.3
43 49.41 $0.54 1M
8
Qwen3.8 Max (0902)
AAlibaba
45.4
— 54.95 $3.00 984K
9
GLM-5.3 (max)
ZZhipu AI
44.8
59 37.15 $2.15 1M
10
Grok 4.6 (high)
XxAI
44.3
64 42.07 $3.00 —
11
Step 5 Preview
43.7
81 27.43 $1.43 1M
12
Kimi K3 (max)
MMoonshot
43.6
— 60.43 $6.00 1M
13
GLM-5.3-Flash
ZZhipu AI
41.8
43 50.64 $0.24 1M
14
Gemini 3.8 Flash (high)
GGoogle
40.9
293 16.03 $1.50 1M
15
DeepSeek V4.1 Flash (max)
DDeepSeek
39.5
232 9.58 $0.53 1M
16
GPT-6 Luna (max)
OOpenAI
37.3
134 125.94 $0.20 1M
17
Qwen3.8 27B (xhigh)
AAlibaba
33.7
43 50.81 $1.13 —
18
K2 Horizon 375B A23B
30.5
— — — 524K
19
MiniMax-M3
MMiniMax
29.2
145 15.15 $0.53 1M
20
Inkling
25
166 13.90 $1.76 1M
Updated: 25 September 2026 · 00:15 Data from Artificial Analysis artificialanalysis.ai ↗