Find a model

Search models and open their dedicated profile. Use the arrow keys to navigate and Enter to open.

AI model leaderboard

Compare intelligence, thinking and cost.

DATA CURRENTUpdated Sep 30, 02:00 AM UTC

Intelligence vs. cost

Each point is a model configuration. Click to explore it.

Scroll the chart ↔ · tap a point for details
01224364860$0$2$3$5$7$9Intelligence Index ↑USD / task AA →↖ more quality, less costClaude Opus 5.5 · maxClaude Opus 5.5 · xhighClaude Sonnet 5.5 · maxClaude Opus 5.5 · highClaude Fable 5.1 · maxClaude Fable 5.1 · xhighGPT-6 Astra · maxGPT-6 Astra · xhighClaude Sonnet 5.5 · xhighGPT-6.1 Sol · maxClaude Opus 5.5 · mediumClaude Fable 5.1 · highGPT-6.1 Sol · xhighGPT-6 Astra · highGPT-6.1 Sol · highGPT-6 Astra · mediumClaude Fable 5.1 · mediumMuse Spark 1.3 · maxGPT-6.1 Sol · mediumGPT-6 Sol · maxClaude Fable 5.1 · lowClaude Sonnet 5.5 · highGrok 4.7 · xhighGrok 4.7 · highMiMo-V2.6-Pro · thinkingGPT-6 Astra · lowQwen3.8 Max (0902) · maxMuse Spark 1.3 · xhighGLM-5.3 · maxGPT-6 Sol · xhighStep 5 Preview · thinkingKimi K3 · maxGPT-6 Sol · highClaude Opus 5.5 · lowGPT-6.1 Sol · lowGPT-5.6 Terra · maxGLM-5.3-Flash · thinkingGemini 3.8 Flash · highClaude Sonnet 5.5 · mediumQwen3.8 2.4T A95B · thinkingQwen3.8-Flash-Next · thinkingGPT-6 Sol · mediumGemini 3.8 Flash · mediumDeepSeek V4.1 Flash · maxGPT-5.6 Terra · xhighMiMo-V2.6-Flash · thinkingGPT-6 Luna · maxDeepSeek V4 Pro 0813 · maxDeepSeek V4 Flash Vision · maxGLM-5.3 · lowGPT-5.6 Terra · highGPT-6 Sol · lowGPT-6 Luna · xhighQwen3.8 27B · xhighGPT-6 Luna · highGPT-5.6 Terra · mediumGemini 3.1 Pro Preview · thinkingGPT-6 Luna · mediumMiniMax-M3 · thinkingGPT-6 Sol · noneQwen3.8 27B · mediumGPT-5.6 Terra · lowQuasar 438B · maxApodex 1.1 · thinkingQwen3.8 27B · lowGPT-5.5 Instant (June 2026) · thinkingKimi K2.7 Code · thinkingHy3 · thinkingQwen3.7 Plus · thinkingDeepSeek V4.1 Flash · noneNemotron 3 Ultra · thinkingGemini 3.5 Flash-Lite · thinkingGPT-6 Luna · lowGPT-5.6 Terra · noneQwen3.8 27B · noneLongCat 2.0 · thinkingQwen3.5 397B A17B · thinkingGPT-6 Luna · noneQwen3.6 35B A3B · thinkingMuse Glimmer · highClaude 4.5 Haiku · thinkingRing-2.6-1T · thinkingQwen3.5 122B A10B · thinkingMistral Medium 3.5 · mediumCommand A+ · thinkingNemotron 3.5 Lightning · thinkingNemotron 3 Super · thinkingMercury 2.5 · thinkinggpt-oss-120b · highMistral Small 4 · thinkingQwen3.5 9B · thinkingGranite 4.2 8B · thinkingLing 3.0 Tiny · thinkingTrinity Large Thinking · thinkingNorth Mini Code · thinkingMistral Large 3 · defaultQwen3 Coder Next · defaultGranite 4.2 3B · thinkinggpt-oss-20b · highNemotron 3 Nano · thinkingCeleris-1 · defaultMinistral 3 14B · defaultMinistral 3 8B · defaultMinistral 3 3B · defaultClaude Opus 5.5Claude Sonnet 5.5Claude Fable 5.1GPT-6 AstraGPT-6.1 SolMuse Spark 1.3GPT-6 Sol
Efficient frontier: best measured score at each cost104 measured configurations · 1813 without both values
AnthropicOpenAIMetaSpaceXAIXiaomiAlibabaZ AIStepFunKimi
Artificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.

Best under budget

GPT-6.1 SolThinking xhigh · $0.3951.0

Fastest with a strong score ≥75% of the highest score

Muse Spark 1.3Thinking max174 tok/s

The full ranking

1,917 configurations · Intelligence Index

#ThinkingCompare
1max57.6$4.00Anthropic$20.0092 tok/s1M
2xhigh56.0$4.00Anthropic$20.0080 tok/s1M
3max56.0$2.00Anthropic$10.00138 tok/s1M
4high53.6$4.00Anthropic$20.0074 tok/s1M
5max53.4$10.00Anthropic$50.0069 tok/s1M
6xhigh53.2$10.00Anthropic$50.0059 tok/s1M
7max52.7$10.00OpenAI$50.0056 tok/s1M
8xhigh52.4$10.00OpenAI$50.0049 tok/s1M
9xhigh51.9$2.00Anthropic$10.00103 tok/s1M
10max51.8$2.00OpenAI$10.0069 tok/s1M
11medium51.2$4.00Anthropic$20.0074 tok/s1M
12high51.2$10.00Anthropic$50.0052 tok/s1M
13xhigh51.0$2.00OpenAI$10.0063 tok/s1M
14high50.9$10.00OpenAI$50.0047 tok/s1M
15high50.2$2.00OpenAI$10.0065 tok/s1M
16medium49.6$10.00OpenAI$50.0046 tok/s1M
17medium48.9$10.00Anthropic$50.0050 tok/s1M
18max48.1$1.25Meta$4.25174 tok/s1M
19medium47.8$2.00OpenAI$10.0059 tok/s1M
20max47.5$2.00OpenAI$10.0076 tok/s872K
21low46.8$10.00Anthropic$50.0048 tok/s1M
22high46.7$2.00Anthropic$10.0090 tok/s1M
23xhigh46.4$2.00SpaceXAI$6.0074 tok/s500K
24high46.3$2.00SpaceXAI$6.0077 tok/s500K
25thinking46.3$0.44Xiaomi$0.8741 tok/s1M

A benchmark is a starting point. Token prices and measured cost per task answer different questions. Check provider, thinking level and source date before you choose.