Updated 4h ago

Models, ranked across every board

59 models earn points on the language-model boards. Podium score is points over points available.

All Open weights Closed

#	Model	Podium score	#1s	Top 10s	Median rank
1	Claude Fable 5.1	57.3	1	11	2
2	GPT-6 Astra	48.0	4	8	3
3	Claude Opus 5	38.7	2	10	4
4	Claude Fable 5	38.7	2	8	4.5
5	Claude Opus 5.5	25.3	2	4	1.5
6	GPT-5.6 Sol	22.7	1	6	7.5
7	Claude Opus 4.6	22.7		5	16.5
8	Claude Opus 4.7	20.7		5	11
9	Muse Spark 1.3	18.0		5	6
10	GPT-5.5	15.3		5	9.5
11	Muse Spark	14.7	1	4	10
12	GPT-5.4 Pro	14.7		4	5
13	Gemini 3.8 Flash	14.7		4	11
14	Gemini 3.1 Pro Preview	14.7		3	17
15	Muse Spark 1.1	14.0	1	3	16.5
16	Gemini 3 Pro	12.7		4	18
17	Qwen3.8 Max (0902)	11.3		3	12.5
18	Kimi K3	9.3		3	12
19	GPT-6 Sol	8.0		2	5
20	Gemini 3.7 Flash	8.0		3	10.5
21	GLM-5.3	8.0		3	15
22	Muse Spark 1.2	7.3		2	17
23	Claude Opus 4.5	6.7	1	1	26
24	GPT-5.6 Terra	6.7		3	13
25	Claude Sonnet 4.5	6.0		1	28
26	GPT-5.5 Pro	5.3		2	7
27	Qwen3.8 2.4T A95B	4.7		1	10
28	Claude Sonnet 5	4.7		1	23.5
29	Claude Sonnet 4.6	4.7		2	27.5
30	GLM 4.6	4.7		1	84
31	Grok 4.7	4.0		2	10
32	Grok 4 1 Fast	4.0		1	70
33	Claude Opus 4.8	3.3		3	16.5
33	DeepSeek V4.1 Flash	3.3		1	16.5
35	Gemini 3.5 Flash	3.3		1	17
36	Grok 4.6	3.3		2	18
37	Ernie 5.1	3.3		1	21
38	Gemini 3.6 Flash	3.3		1	22.5
39	GPT-5.1	3.3		1	38
40	Claude 4.5 Haiku	3.3		1	55
41	GPT-5.5 Pre Release	2.7		1	7
41	MiMo-V2.6-Pro	2.7		1	7
43	GPT-5	2.7		1	49
44	o3	2.7		1	52.5
45	GPT-5.4	2.0		1	16
46	Grok 4.5	2.0		1	26
47	o3-pro	2.0		1	33
48	Grok 4	2.0		1	45
49	Gemini 3	1.3		1	9
49	Gemini 3.1 Pro Grounding	1.3		1	9
49	GPT-5.5 Pro Pre Release	1.3		1	9
49	Moonshotai Kimi K2	1.3		1	9
53	Qwen3.8-Flash-Next	1.3		1	13
54	Kimi K2.5	1.3		1	41
55	Gemini 3 Pro Grounding	0.7		1	10
55	Grok 4 1 Fast Non	0.7		1	10
55	Smaug Agentic	0.7		1	10
58	Inkling	0.7		1	45
59	Gemini 3.1 Flash-Lite	0.7		1	59

Leaders elsewhere

Top 10 on at least one board outside the language-model set: image, video, speech, embeddings, speed, price, usage.

GLM 5.3 Flash10 boards 1Agnes 3.0 Flash2 boards 1

Gemini Omni 1.1 Flash2 boards 1

GPT Image 2.5 Sunburst2 boards 1

MiniMax H32 boards 1

harrier-oss-v1-27b (microsoft)1 board 1

Sonic 3.61 board 1scribe v2 pro (zoom)1 board 2

Gemini 3 Flash10 boards 2

MiniMax-M2.55 boards 2Agnes 2.5 Pro Beta2 boards 2

Gemini Omni Flash2 boards 2

GPT Image 2.5 Flare2 boards 2

Gemini 3.8 Flash TTS1 board 2

azure-speech-07-2026 (microsoft)1 board 2

KaLM-Embedding-Gemma3-12B-2511 (tencent)1 board 3

Gemini 2.5 Pro (Jun 2025)12 boards 3

GPT-6 Luna5 boards 3

Flux 3 Video2 boards 3

GPT Image 22 boards 3

Hy4 preview2 boards 3

Wan3.02 boards 3multilingual (modulate)1 board 3

llama-embed-nemotron-8b (nvidia)1 board 3

Qwen Audio 3.0 TTS Plus1 board 4

DeepSeek V4 Flash Vision10 boards 4

GPT-5.6 Luna9 boards 4

Dreamina Seedance 2.5 720p2 boards 4

Grok Imagine Image 2.02 boards 4

MAI Image 2.62 boards 4

scribe v2 (elevenlabs)1 board 4

Grok Imagine Video 1.5 Agent1 board 4

Qwen3-Embedding-8B (Qwen)1 board 4

Realtime TTS 21 board 5

GPT-5.3 Codex5 boards 5

Dreamina Seedance 2.0 720p2 boards 5

gemini-embedding-001 (google)1 board 5resonant-1 (reson8)1 board 5Simba 3.21 board 6

GPT-5.214 boards 6

GLM-57 boards 6

MiMo-V2.55 boards 6

GPT-5.2 Codex3 boards 6

MiniMax M2.12 boards 6

Muse Image2 boards 6Reve 2.12 boards 6

Gemini 3.8 Flash Lite TTS1 board 6

o3 GPT 4.11 board 6

Qwen3-Embedding-4B (Qwen)1 board 6resonant-1-flash (reson8)1 board 7

Qwen3.5 397B A17B7 boards 7

Nemotron 3 Ultra 550B A55B3 boards 7

MAI Image 2.52 boards 7

Gemini 2.5 Pro 05 061 board 7Hidream o1 Video 1.01 board 7Luna TTS1 board 7Octen-Embedding-8B (Octen)1 board 7scribe v1 (zoom)1 board 8

DeepSeek-V3.28 boards

Scores as published by each board on the capture date. Model names and logos belong to their owners; logos via logo.dev.

Weekly: the AI leaderboards with a new #1, Saturday mornings.