Skip to content

Benchmarks

Sources
8
Entries tracked
22,318
Last sync
30 min
All sources syncing normally
  • Arena · Text2 hours ago
  • Arena · WebDev2 hours ago
  • Arena · Vision2 hours ago
  • Arena · Agent2 hours ago
  • GPQA Diamond30 minutes ago
  • SWE-bench30 minutes ago
  • Terminal-Bench30 minutes ago
  • ARC-AGI-230 minutes ago
Chinese ×1372 of 372 models
ModelOrg± CISwingAdd to compare
1Gemini 4 Argon (High)googleOct 2026$2 / $101,5871,555–1,618#1–10
2Claude Opus 5.5 (High)anthropicOct 2026$4 / $201,5721,551–1,594#2–11
3Claude Fable 5.1 (Max)anthropicSep 2026$10 / $501,5691,550–1,587#2–22
4Claude Fable 5 (High)anthropicSep 2026$10 / $501,5531,542–1,564#4–21
5Claude Opus 4.6 (High)anthropicSep 2026$5 / $251,5501,541–1,559#2–25
6Claude Opus 5 (High)anthropicSep 2026$5 / $251,5471,537–1,556#5–11
7claude-sonnet-5.5-xhighanthropicOct 2026$2 / $101,5451,520–1,570#4–26
8Gemini 3.7 Flash (High)googleOct 2026$0.75 / $3.751,5441,531–1,558#9–57
9Muse Wonder (Max)metaOct 2026—1,5431,507–1,578#14–56
10gpt-5.6-sol-xhighopenaiSep 2026$4 / $201,5411,530–1,553#6–24
11Kimi K3 (Max)moonshotJun 2026$0.24 / $11.941,5411,528–1,553#10–40
12Claude Opus 4.7 (High)anthropicSep 2026$5 / $251,5381,528–1,548#1–18
13Gemini 3.8 Flash (High)googleOct 2026$0.83 / $4.131,5361,525–1,548#2–30
14qwen3.8-maxalibabaOct 2026$2 / $61,5351,522–1,549#12–41
15glm-5.3-maxzaiAug 2026$0.08 / $1.401,5351,520–1,550#18–53
16Muse Spark 1.3 (Max)metaSep 2026$1.25 / $4.251,5301,514–1,547#13–33
17gemini-3.6-flash-highgoogleSep 2026$0.75 / $3.751,5301,518–1,542#16–95
18Muse Spark 1.1metaSep 2026$1.25 / $4.251,5291,518–1,540#21–115
19glm-5.3-flashzaiAug 2026$0.05 / $0.201,5291,516–1,541#26–59
20Gemini 3.1 Pro PreviewgoogleSep 2026$2 / $121,5271,520–1,535#9–36
21claude-opus-4-8-highanthropicSep 2026$5 / $251,5261,516–1,536#15–33
22gpt-5.5openaiSep 2026$5 / $301,5251,515–1,535#9–34
23Muse SparkmetaSep 2026—1,5251,505–1,545#16–41
24deepseek-v4.1-flash-maxdeepseekOct 2026$0.15 / $0.601,5251,506–1,544#14–30
25qwen3.7-max-previewalibabaOct 2026—1,5231,488–1,558#29–73
26gpt-6.1-sol-maxopenaiOct 2026$2 / $101,5231,498–1,548#1–15
27glm-5.2-maxzaiJun 2026$0.62 / $2.421,5211,510–1,531#9–49
28Gemini 3 ProgoogleJan 2026—1,5211,510–1,531#23–91
29gemini-3.5-flash-mediumgoogleSep 2026$1.50 / $91,5181,507–1,528#10–26
30gemini-3-flashgoogleJan 2026—1,5171,505–1,529#38–153
31kimi-k2.6moonshotApr 2026$0.65 / $3.411,5171,504–1,529#30–77
32gpt-5.6-terra-xhighopenaiSep 2026$2 / $121,5171,506–1,527#13–60
33gpt-6-astra-maxopenaiSep 2026$10 / $501,5161,497–1,535#1–18
34qwen3.5-max-previewalibabaOct 2026—1,5161,500–1,531#33–77
35glm-5zaiFeb 2026$1 / $3.201,5161,502–1,529#57–153

Showing Chinese exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.

The few things in AI and quantum that actually mattered this week — honestly summarised.

One email a week. Unsubscribe in one click. No tracking, no sharing.