Skip to content
IFAI

Model rankings

443 models · 2 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Text arena · Anonymous head-to-head votes · Published July 25, 2025

Most preferred
gemini-2.5-proGoogle1,460Score
Best open model
kimi-k2-0711-previewMoonshot AI1,418Score
Best from China
kimi-k2-0711-previewMoonshot AI1,418Score
Most voted
claude-3-opus-20240229Anthropic202,641Votes
Top three by Score
1gemini-2.5-proGoogle1,460
2o3-2025-04-16OpenAI1,452
Top three open models
1kimi-k2-0711-previewMoonshot AI · #6 overall1,418
2deepseek-r1-0528DeepSeek · #8 overall1,416
3qwen3-235b-a22b-no-thinkingAlibaba · #14 overall1,398
This week's changes
Kimi K2Intelligence · #22New
Claude 4 Sonnet (Extended Thinking)Intelligence · #13Up 4
RankChangeModelScore95% intervalVotesPrice $/M
1—gemini-2.5-proClosedGoogle1,460±622,146—
2—o3-2025-04-16ClosedOpenAI1,452±528,218—
3—chatgpt-4o-latest-20250326ClosedOpenAI1,443±528,115—
4—gpt-4.5-preview-2025-02-27ClosedOpenAI1,438±615,271—
5—grok-4-0709ClosedxAI1,432±79,270—
6—claude-opus-4-20250514-thinking-16kClosedAnthropic1,417±615,388—
7—claude-opus-4-20250514ClosedAnthropic1,413±523,514—
8▲ 2grok-3-preview-02-24ClosedxAI1,410±529,333—
9▼ 1gemini-2.5-flashClosedGoogle1,410±527,556—
10▼ 1gpt-4.1-2025-04-14ClosedOpenAI1,408±522,106—
11—claude-sonnet-4-20250514-thinking-32kClosedAnthropic1,399±614,416—
12▲ 1o1-2024-12-17ClosedOpenAI1,398±429,038—
13▼ 1o4-mini-2025-04-16ClosedOpenAI1,398±521,868—
14—claude-sonnet-4-20250514ClosedAnthropic1,388±619,867—
15▲ 1o1-previewClosedOpenAI1,385±533,177—
16▼ 1claude-3-7-sonnet-20250219-thinking-32kClosedAnthropic1,384±529,944—
17—gpt-4.1-mini-2025-04-14ClosedOpenAI1,373±521,243—
18—claude-3-7-sonnet-20250219ClosedAnthropic1,370±434,847—
19▲ 3gemini-2.0-flash-001ClosedGoogle1,366±439,017—
20▼ 1claude-3-5-sonnet-20241022ClosedAnthropic · 1 more variant1,365±381,824—
21▼ 1o3-mini-highClosedOpenAI1,363±519,404—
22▼ 1gemma-3-27b-itOpenGoogle1,362±529,689—
23—grok-3-mini-highClosedxAI1,356±77,711—
24—grok-3-mini-betaClosedxAI1,354±614,313—
25—gemini-2.0-flash-lite-preview-02-05ClosedGoogle1,351±426,104—
26—gemini-1.5-pro-002ClosedGoogle1,350±358,645—
27▲ 1llama-3.1-nemotron-ultra-253b-v1OpenNVIDIA1,344±122,656—
28▼ 1o3-miniClosedOpenAI1,344±440,563—
29—gpt-4o-2024-05-13ClosedOpenAI1,343±3117,747—
30—gemma-3-12b-itOpenGoogle1,340±93,976—
31▲ 1llama-3.1-405b-instruct-bf16OpenMeta1,334±443,788—
32▼ 1o1-miniClosedOpenAI1,334±454,951—
33▲ 1llama-3.1-405b-instruct-fp8OpenMeta1,333±363,038—
34▲ 2gpt-4o-2024-08-06ClosedOpenAI1,332±447,973—
35▲ 2grok-2-2024-08-13ClosedxAI1,331±367,084—
36▼ 3gemini-advanced-0514ClosedGoogle1,331±552,144—
37▼ 2llama-4-maverick-17b-128e-instructOpenMeta1,330±521,923—
38—llama-3.3-nemotron-49b-super-v1OpenNVIDIA1,324±122,371—
39—gpt-4-turbo-2024-04-09ClosedOpenAI1,323±4102,133—
40▲ 3claude-3-opus-20240229ClosedAnthropic1,321±3202,641—
41▲ 1gemini-1.5-pro-001ClosedGoogle1,320±482,435—
42▼ 1gpt-4.1-nano-2025-04-14ClosedOpenAI1,320±86,302—
43▼ 3llama-4-scout-17b-16e-instructOpenMeta1,316±711,604—
44▲ 1llama-3.3-70b-instructOpenMeta1,316±449,882—
45▼ 1claude-3-5-haiku-20241022ClosedAnthropic1,315±453,297—
46▲ 2gpt-4o-mini-2024-07-18ClosedOpenAI1,315±372,423—
47—gpt-4-1106-previewClosedOpenAI1,314±4103,748—
48▲ 1gpt-4-0125-previewClosedOpenAI1,314±497,079—
49▼ 3gemma-3n-e4b-itOpenGoogle1,313±710,827—
50—gemini-1.5-flash-002ClosedGoogle1,312±437,021—

Intelligence over time

Each maker's best model at the time on the Artificial Analysis Intelligence Index, by the month it was released.

Gained most · Anthropic +57.2Highest now · Anthropic 64.6Gap between highest and lowest 18.1 ptsTime range Nov 2025 – Nov 2026
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi

Arena data comes from the leaderboard dataset it publishes on Hugging Face (CC BY 4.0). Only each board's overall leaderboard is used; scores and ranks are unchanged. dataset · CC BY 4.0

Artificial Analysis data comes from its API (artificialanalysis.ai). Scores are unchanged; ranks follow the scores. artificialanalysis.ai

Open or closed follows the license Arena gives; prices and Artificial Analysis intelligence show only where model names match, otherwise “—”.