Skip to content
IFAI

Model rankings

1125 models · 5 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Text arena · Anonymous head-to-head votes · Published April 10, 2026

Most preferred
claude-opus-4-6-thinkingAnthropic1,504Score
Best open model
glm-5.1Z.ai1,471Score
Best from China
glm-5.1Z.ai1,471Score
Most voted
claude-3-opus-20240229Anthropic194,909Votes
Top three by Score
1claude-opus-4-6-thinkingAnthropic1,504
2muse-sparkMeta1,493
Top three open models
1glm-5.1Z.ai · #12 overall1,471
2glm-5Z.ai · #21 overall1,455
3kimi-k2.5-thinkingMoonshot AI · #24 overall1,452
This week's changes
HappyHorse-1.0Video · #1New
glm-5.1Coding · #2New
Eleven v3Speech · #2New
Speech 2.8 HDSpeech · #4New
RankChangeModelScore95% intervalVotesPrice $/M
1—claude-opus-4-6-thinkingClosedAnthropic · 1 more variant1,504±516,278—
2Newmuse-sparkClosedMeta1,493±103,268—
3▼ 1gemini-3.1-pro-previewClosedGoogle1,492±520,531—
4—gemini-3-proClosedGoogle1,486±441,585—
5▼ 2grok-4.20-beta1ClosedxAI1,486±79,689—
6▼ 1grok-4.20-beta-0309-reasoningClosedxAI1,478±69,781—
7▼ 1gpt-5.2-chat-latest-20260210ClosedOpenAI1,477±515,704—
8—grok-4.20-multi-agent-beta-0309ClosedxAI1,476±610,112—
9▼ 2gemini-3-flashClosedGoogle · 1 more variant1,474±430,918—
10▼ 1claude-opus-4-5-20251101-thinking-32kClosedAnthropic1,473±437,307—
11▼ 1grok-4.1-thinkingClosedxAI · 1 more variant1,471±447,508—
12▼ 1claude-opus-4-5-20251101ClosedAnthropic1,468±447,320—
13▼ 1qwen3.5-max-previewClosedAlibaba1,466±77,952—
14▼ 1gpt-5.4ClosedOpenAI1,466±79,977—
15—claude-sonnet-4-6ClosedAnthropic1,462±610,940—
16▼ 2dola-seed-2.0-proClosed1,461±518,882—
17—gpt-5.4-mini-highClosedOpenAI1,459±77,169—
18▼ 2gpt-5.3-chat-latestClosedOpenAI1,456±614,444—
19▼ 1gpt-5.1-highClosedOpenAI1,454±441,042—
20▼ 1claude-sonnet-4-5-20250929-thinking-32kClosedAnthropic1,452±360,401—
21▼ 1claude-sonnet-4-5-20250929ClosedAnthropic1,451±358,292—
22▼ 1ernie-5.0-0110ClosedBaidu1,450±522,778—
23▼ 1ernie-5.0-preview-1203ClosedBaidu · 1 more variant1,449±79,810—
24▼ 1claude-opus-4-1-20250805-thinking-16kClosedAnthropic1,448±350,152—
25▼ 1gemini-2.5-proClosedGoogle1,448±3107,824—
26▼ 1claude-opus-4-1-20250805ClosedAnthropic1,447±377,864—
27▼ 1mimo-v2-proClosedXiaomi1,446±78,397—
28—gpt-4.5-preview-2025-02-27ClosedOpenAI1,444±614,547—
29—chatgpt-4o-latest-20250326ClosedOpenAI1,443±382,998—
30—gpt-5.2-highClosedOpenAI1,442±430,488—
31▼ 4longcat-flash-chat-2602-expClosedMeituan1,440±85,790—
32▼ 1gpt-5.2ClosedOpenAI1,439±427,564—
33▼ 1gpt-5.1ClosedOpenAI1,439±443,708—
34▼ 1gemini-3.1-flash-lite-previewClosedGoogle1,435±515,996$0.56
35▼ 1qwen3-max-previewClosedAlibaba1,435±427,940—
36▼ 1gpt-5-highClosedOpenAI1,433±532,259—
37▼ 1grok-4-1-fast-reasoningClosedxAI1,432±442,592—
38▼ 1o3-2025-04-16ClosedOpenAI1,431±460,172—
39▼ 1amazon-nova-experimental-chat-26-02-10ClosedAmazon1,428±103,452—
40▼ 1gpt-5-chatClosedOpenAI1,426±431,851—
41▼ 1qwen3-max-2025-09-23ClosedAlibaba1,424±69,242—
42▼ 1claude-opus-4-20250514-thinking-16kClosedAnthropic1,424±437,191—
43▼ 1grok-4-fast-chatClosedxAI1,421±86,872—
44▼ 1amazon-nova-experimental-chat-26-01-10ClosedAmazon1,415±103,436—
45▼ 1gpt-4.1-2025-04-14ClosedOpenAI1,413±451,411—
46▼ 1claude-opus-4-20250514ClosedAnthropic1,412±444,557—
47▼ 1grok-3-preview-02-24ClosedxAI1,412±433,053—
48▼ 1gemini-2.5-flashClosedGoogle1,411±3107,221—
49—grok-4-0709ClosedxAI1,410±441,742—
50▼ 2mistral-medium-2508ClosedMistral · 2 more variants1,410±377,340—

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 “—”.