Independent evaluations
AI model leaderboard
Only third-party results, measured the same way for every model. Vendor-reported scores are never counted.
- Evaluations
- 6
- Ranked models
- 60
- Updated
- 15 Nov, 00:20
- Runner-up
Gemini 4 Carbon
Google · Closed weights
60.0Intelligence IndexFour-area profile- Context
- 1.05M
- Blended price
- $4.00
- Released
- 10.28
- Champion
Claude Fable 5.2
Anthropic · Closed weights
64.6Intelligence IndexFour-area profile- Context
- 1M
- Blended price
- $16.0
- Released
- 11.09
- Third place
Claude Opus 5.5
Anthropic · Closed weights
59.1Intelligence IndexFour-area profile- Context
- 1M
- Blended price
- $8.00
- Released
- 09.22
| # | Model | By area | Cached input | ||||||
|---|---|---|---|---|---|---|---|---|---|
| 01 | Anthropic | 64.6 | 100% | $8.00 | $40.0 | $0.20 | 91 | 2026-11-09 | |
| 02 | Google | 60.0 | 100% | $2.00 | $10.0 | $0.10 | 98 | 2026-10-28 | |
| 03 | Anthropic | 59.1 | 100% | $4.00 | $20.0 | $0.20 | 97 | 2026-09-22 | |
| 04 | OpenAI | 56.9 | 100% | $10.0 | $50.0 | $1.00 | 50 | 2026-09-03 | |
| 05 | OpenAI | 54.6 | 100% | $2.00 | $10.0 | $0.10 | 64 | 2026-09-29 | |
| 06 | Anthropic | 54.5 | 100% | $2.00 | $10.0 | $0.10 | 135 | 2026-09-28 | |
| 07 | Google | 53.9 | 100% | $1.50 | $7.50 | $0.075 | 232 | 2026-11-12 | |
| 08 | OpenAI | 51.3 | 100% | $1.00 | $5.00 | $0.050 | 85 | 2026-10-26 | |
| 09 | Google | 50.4 | 100% | $2.00 | $10.0 | $0.10 | 86 | 2026-09-30 | |
| 10 | Meta | 50.1 | 100% | Not published | Not published | Not published | — | 2026-10-12 | |
| 11 | Anthropic | 50.0 | 100% | $10.0 | $50.0 | $0.25 | 59 | 2026-09-01 | |
| 12 | KimiOpen weights | 46.5 | 100% | $2.00 | $10.0 | $0.20 | 29 | 2026-10-23 | |
| 13 | Anthropic | 44.2 | 100% | $5.00 | $25.0 | $0.50 | 54 | 2026-07-24 | |
| 14 | OpenAI | 44.2 | 100% | $2.00 | $10.0 | $0.20 | 150 | 2026-09-22 | |
| 15 | Meta | 43.6 | 100% | $4.00 | $20.0 | $0.20 | 151 | 2026-10-12 | |
| 16 | Meta | 36.6 | 100% | $1.25 | $4.25 | $0.15 | 145 | 2026-09-02 | |
| 17 | Anthropic | 36.5 | 100% | $0.10 | $0.50 | $0.010 | 235 | 2026-10-07 | |
| 18 | Meta | 36.0 | 100% | $0.10 | $0.50 | $0.004 | 647 | 2026-11-09 | |
| 19 | SpaceXAI | 35.5 | 100% | $2.00 | $6.00 | $0.50 | 66 | 2026-09-21 | |
| 20 | Tencent | 34.6 | 100% | $0.83 | $2.50 | $0.042 | 95 | 2026-11-03 | |
| 21 | Alibaba | 33.7 | 100% | $2.00 | $6.00 | $0.25 | 39 | 2026-09-02 | |
| 22 | DeepSeekOpen weights | 32.6 | 100% | $0.30 | $1.20 | $0.006 | 173 | 2026-10-17 | |
| 23 | OpenAI | 32.3 | 100% | $2.00 | $12.0 | $0.20 | 123 | 2026-07-09 | |
| 24 | Z AIOpen weights | 31.7 | 100% | $1.40 | $4.40 | $0.26 | 94 | 2026-08-18 | |
| 25 | SpaceXAI | 31.0 | 100% | $2.00 | $6.00 | $0.50 | 83 | 2026-08-12 | |
| 26 | Google | 30.8 | 100% | $0.75 | $3.75 | $0.075 | 126 | 2026-09-02 | |
| 27 | Thinking MachinesOpen weights | 30.6 | 100% | $0.30 | $1.00 | $0.030 | 77 | 2026-11-05 | |
| 28 | XiaomiOpen weights | 29.9 | 100% | $0.43 | $0.87 | $0.004 | 62 | 2026-09-21 | |
| 29 | InclusionAIOpen weights | 29.6 | 100% | $0.30 | $0.90 | $0.060 | — | 2026-10-22 | |
| 30 | MiniMaxOpen weights | 29.1 | 100% | $0.15 | $0.60 | $0.015 | 159 | 2026-10-19 |
New models last 30 days
- Gemini 4 HydrogenGoogle · 2026-11-12#7
- JT-4.1 Flash 236B A21BChina Mobile · 2026-11-11#44
- Claude Fable 5.2Anthropic · 2026-11-09#1
- Muse EmberMeta · 2026-11-09#18
- Agnes 2.5 Pro ThetaSapiens AI · 2026-11-07#39
- Inkling 2 SmallThinking Machines · 2026-11-05#27
- Tencent Hy4Tencent · 2026-11-03#20
- Mistral Large 4Mistral · 2026-10-29#36
- Gemini 4 CarbonGoogle · 2026-10-28#2
- GPT-6.1 TerraOpenAI · 2026-10-26#8
- Nanbeige4.3-33BNanbeige · 2026-10-25#42
- Kimi K3.1 ProKimi · 2026-10-23#12
Not yet ranked
Measured in some areas, but Artificial Analysis has not published an Intelligence Index for them yet. They are ranked automatically once it does.
- Qwen3.8 2.4T A95B5 evaluations · missing: Engineering—
- GLM-5.25 evaluations · missing: Engineering—
- Granite 4.2 30B4 evaluations · missing: Engineering, Healthcare—
- Solar Mini 41 evaluations · missing: Engineering, Legal, Healthcare—
- Ling 3.0 Tiny1 evaluations · missing: Engineering, Legal, Healthcare—
- A.X-K21 evaluations · missing: Engineering, Legal, Healthcare—
- Solar Pro 31 evaluations · missing: Engineering, Legal, Healthcare—
- Ring-2.6-1T1 evaluations · missing: Engineering, Legal, Healthcare—
How to read this board
Every number is an index published by Artificial Analysis on a 0–100 scale, shown unchanged. The ranking follows the Artificial Analysis Intelligence Index; models released in the last 12 months are ranked.
Area scores: engineering, legal and healthcare use their own index; business is the average of finance & accounting, strategy & operations and economics. A missing index is neither filled in nor penalised.
Prices are list prices per million tokens and do not affect the ranking.
Data: Artificial Analysis · OpenRouter