Ranked AI Models

The overall score ranks large language models from 0 to 100. It averages the published scores for knowledge freshness, release recency, context window, parameter count, and token price. A model is included when at least three of those scores exist, and a missing score is left out of the average. A higher score is better on those measures. See the highest ranks.

Data sourced from OpenRouter API and manual data collection | Last updated: Nov 4, 2026, 8:10 AM UTC

Ranked by overall score

80–100 overall score

Ranked by overall score, 80–100
Rank	Model	Overall	Freshness	Recency	Context	Parameters	Price
1	Z.ai: GLM 5.3 Flash	89	—	97	71	—	100
2	Qwen: Qwen3.7 Flash	88	—	95	70	—	99
3	OpenAI: GPT-6 Luna	86	92	98	71	—	82
4	Particle AI: Particle-1	85	—	100	71	—	84
4	Meta: Muse Spark 1.3 Contributor	85	—	97	71	—	88
4	Meta: Muse Spark 1.2 Contributor	85	—	96	71	—	88
7	Inception: Mercury 3	84	—	99	70	—	84
7	Anthropic: Claude Haiku 5.5	84	—	99	70	—	82
7	OpenAI: GPT-6 Luna Pro	84	—	98	71	—	82
10	MiniMax: MiniMax M3.1 Flash	83	—	99	71	—	78
10	Qwen: Qwen3.8 Omni Flash	83	—	98	70	—	80
10	Nex AGI: Nex-N2.5-Mini	83	—	97	53	—	100
13	inclusionAI: Ling 3.1 Flash	82	—	98	70	69	89
13	Qwen: Qwen3.8 Flash	82	—	97	70	—	80
13	Inception: Mercury 2.5	82	—	97	53	—	96
13	Qwen: Qwen3.5-Flash	82	—	88	70	—	89
17	OpenAI: GPT-5.6 Luna	81	88	94	71	—	71
17	OpenAI: GPT-5.6 Luna Pro	81	88	94	71	—	71
17	DeepSeek: DeepSeek V4 Flash Vision Exp	81	—	96	71	—	75
20	Reflection: Beam 1.0 501B A23B	80	94	99	71	67	71
20	Z.ai: GLM 5.3 FlashX Legal	80	—	100	71	—	68
20	Nex AGI: Nex-N2.5-Pro	80	—	97	53	—	89

🤖 Generate website with AI 👆

Join the beta waitlist and get 50% off your first 3 months.

Join waitlist

60–79 overall score

Ranked by overall score, 60–79
Rank	Model	Overall	Freshness	Recency	Context	Parameters	Price
23	Tencent: Hy4	79	—	100	71	—	67
23	Xiaomi: MiMo-V2.6-Flash	79	—	98	71	62	84
23	Z.ai: GLM 5.3 FlashX	79	—	98	71	—	68
23	Upstage: Solar Pro 4	79	—	96	62	—	78
23	Meituan: LongCat 2.0	79	—	95	71	80	69
23	DeepSeek: DeepSeek V4 Flash 0731	79	—	95	71	61	89
29	Google: Gemini 3.8 Flash	78	90	97	71	—	54
29	Google: Gemini 3.7 Flash	78	90	96	71	—	55
29	Xiaomi: MiMo-V2.6-Pro	78	—	98	71	75	69
29	MiniMax: MiniMax M3	78	—	93	71	—	69
33	Qwen: Qwen3.7 Plus	77	—	93	70	—	68
34	Anthropic: Claude Sonnet 5.5	76	94	98	70	—	42
34	OpenAI: GPT-6 Sol	76	91	98	71	—	42
34	OpenAI: GPT-6.1 Sol	76	91	98	71	—	42
34	SpaceXAI: Grok 4.3	76	86	91	70	—	55
34	SpaceXAI: Grok 4.20	76	80	89	79	—	55
34	SpaceXAI: Grok 4.20 Multi-Agent	76	80	89	79	—	55
34	Mistral: Mistral Large 4	76	—	99	71	75	60
34	NVIDIA: Nemotron 3.5 Super VL	76	—	99	71	52	80
34	DeepSeek: DeepSeek V4 Pro 0813	76	—	96	71	80	58
34	Z.ai: GLM 5.3	76	—	96	71	—	61
34	Qwen: Qwen3.5 Plus 2026-04-20	76	—	91	70	—	66
45	SpaceXAI: Grok 4.7	75	93	98	62	—	46
45	OpenAI: GPT-5.4 Nano	75	80	89	59	—	71
45	Google: Gemini 3.5 Flash Lite	75	70	95	71	—	63
45	Google: Gemini 3.1 Flash Lite	75	70	91	71	—	68
45	Google: Gemini 2.5 Flash Lite	75	70	—	71	—	84
45	Pareto 26.10: Pareto 26.10	75	—	99	71	—	56
45	Tencent: Hy4 preview	75	—	97	71	72	58
45	Qwen: Qwen3.8 2.4T A95B	75	—	96	71	85	46
45	inclusionAI: Ling 3.0 Flash	75	—	95	53	52	100
45	Xiaomi: MiMo-V2.5-Pro	75	—	91	71	—	63
45	Xiaomi: MiMo-V2.5	75	—	91	70	—	64
45	ByteDance Seed: Seed-2.0-Mini	75	—	88	53	—	84
57	Anthropic: Claude Opus 5.5	74	94	98	70	—	33
57	OpenAI: GPT-5.6 Terra	74	88	94	71	—	41
57	OpenAI: GPT-5.6 Terra Pro	74	88	94	71	—	41
57	Google: Gemini 3.1 Flash Lite Preview	74	70	88	71	—	68
57	OpenAI: GPT-6.1 Terra	74	—	100	71	—	51
57	OpenAI: GPT-6.1 Terra Pro	74	—	100	71	—	51
57	Meta: Muse Spark Open	74	—	99	71	73	51
57	StepFun: Step 5 Preview	74	—	99	71	69	56
57	MoonshotAI: Kimi K3	74	—	95	71	86	42
57	Tencent: Hy3 preview	74	—	91	53	—	77
57	Z.ai: GLM 4.7 Flash	74	—	86	50	—	86
68	Sakana: Fugu Ultra v2	73	97	97	70	—	29
68	SpaceXAI: Grok 4.6	73	87	96	62	—	46
68	Anthropic: Claude Sonnet 5	73	87	94	70	—	42
68	Qwen: Qwen3 Coder Flash	73	77	—	70	—	73
68	Google: Gemini 3.6 Flash	73	70	95	71	—	55
68	Upstage: Solar Mini 4	73	—	98	62	39	93
68	Meta: Muse Spark 1.3	73	—	97	71	—	51
68	Meta: Muse Spark 1.2	73	—	96	71	—	51
68	Z.ai: GLM 5.2	73	—	93	71	—	54
77	Anthropic: Claude Opus 5	72	93	95	70	—	30
77	OpenAI: GPT-5.6 Sol	72	88	94	71	—	33
77	OpenAI: GPT-5.6 Sol Pro	72	88	94	71	—	33
77	SpaceXAI: Grok 4.5	72	87	94	62	—	46
77	Qwen: Qwen-Plus	72	73	70	70	—	73
77	OpenAI: GPT-4.1 Nano	72	60	73	71	—	84
77	Meta: Muse Spark 1.1	72	—	94	71	—	51
77	Tencent: Hy3	72	—	94	53	62	80
77	ByteDance Seed: Seed 1.6 Flash	72	—	77	53	—	87
86	Anthropic: Claude Fable 5.1	71	94	97	70	—	21
86	OpenAI: GPT-5.4 Mini	71	80	89	59	—	54
86	Google: Gemini 3 Flash Preview	71	70	85	71	—	59
86	OpenAI: GPT-5 Nano	71	58	79	59	—	87
86	Google: Gemini 4 Argon	71	—	99	71	—	42
86	MoonshotAI: Kimi K3.1 Pro	71	—	99	71	—	42
86	AionLabs: Aion 3.5 Mini	71	—	98	53	—	62
86	Qwen: Qwen3.8 Max (0902)	71	—	97	70	—	46
86	Sakana: Fugu Max	71	—	97	70	—	46
86	inclusionAI: Ling 3.0 Flash VL	71	—	97	44	52	92
86	Thinking Machines: Inkling	71	—	95	62	75	53
86	Thinking Machines: Inkling Small	71	—	95	62	61	67
86	Qwen: Qwen3.7 Max	71	—	92	70	—	50
99	OpenAI: GPT-6 Astra	70	91	97	71	—	21
99	Amazon: Nova Lite 1.0	70	66	—	55	—	90
99	Elorian: Glimpse	70	—	99	53	—	58
99	OpenAI: GPT-6 Sol Pro	70	—	98	71	—	42
99	OpenAI: GPT-6.1 Sol Pro	70	—	98	71	—	42
99	Z.ai: GLM 5.3 Prime	70	—	98	70	—	41
99	ByteDance Seed: Seed 2.1 Turbo	70	—	96	53	—	61
99	MiniMax: MiniMax M2.5	70	—	88	50	—	71
107	OpenAI: GPT-5.5	69	84	91	71	—	29
107	Google: Gemini 3.5 Flash	69	68	92	71	—	45
107	Amazon: Nova Micro 1.0	69	66	—	44	—	97
107	Meta: Llama 4 Scout	69	63	—	74	51	86
107	Xiaomi: MiMo-V2.6-Pro-UltraSpeed	69	—	98	71	—	38
107	Fireworks: Ember-1	69	—	98	71	—	37
107	NVIDIA: Nemotron 3.5 Lightning	69	—	96	53	37	91
107	ByteDance Seed: Seed-2.0-Code	69	—	95	53	—	59
107	NVIDIA: Nemotron 3 Ultra	69	—	93	53	69	60
107	ByteDance Seed: Seed-2.0-Lite	69	—	89	53	—	66
107	MiniMax: MiniMax M2.7	69	—	89	50	—	69
107	DeepSeek: DeepSeek V3.2	69	—	84	47	—	77
107	OpenAI: GPT-5.1-Codex-Mini	69	—	83	59	—	66
120	Google: Gemini 3.1 Pro Preview	68	70	88	71	—	41
120	Google: Gemini 3.1 Pro Preview Custom Tools	68	70	88	71	—	41
120	Google: Gemini 2.5 Flash	68	70	—	71	—	63
120	Cohere: Command R7B (12-2024)	68	63	—	44	—	96
120	Meta: Muse Wonder	68	—	99	71	—	33
120	Qwen: Qwen3.8 Max Prime	68	—	98	70	—	37
120	PrismML: Ternary Bonsai 2 27B	68	—	98	53	36	83
120	Qwen: Qwen3.8 27B	68	—	96	71	36	68
120	Sakana: Sakana Namazu	68	—	96	53	—	54
120	MoonshotAI: Kimi K2.7 Code	68	—	93	53	—	57
120	StepFun: Step 3.7 Flash	68	—	92	53	57	71
120	SpaceXAI: Grok Build 0.1	68	—	92	53	—	58
120	Inception: Mercury 2	68	—	88	44	—	73
133	DeepSeek: DeepSeek V3.2 Exp	67	78	—	47	—	76
133	Qwen: Qwen3 Coder 480B A35B	67	77	—	53	67	70
133	Z.ai: GLM 4.6	67	73	84	50	—	62
133	OpenAI: GPT-4.1 Mini	67	60	73	71	—	65
133	PrismML: Ternary Bonsai 2 9B	67	—	99	53	24	92
133	AionLabs: Aion-3.0-Mini	67	—	94	44	—	62
133	MoonshotAI: Kimi K2.6	67	—	91	53	—	57
140	Qwen: Qwen3 235B A22B Instruct 2507	66	77	—	53	59	76
140	Mistral: Codestral 2508	66	73	—	53	—	71
140	DeepSeek: DeepSeek V3.1	66	73	—	47	71	72
140	OpenAI: GPT-5 Mini	66	58	79	59	—	66
140	MiniMax: MiniMax-01	66	56	—	70	66	72
140	DeepSeek: DeepSeek V4.1 Flash 1017	66	—	99	71	23	69
140	Google: Gemma 4 26B A4B	66	—	90	53	35	86
140	OpenAI: GPT-5.4	66	—	89	71	—	38
140	Qwen: Qwen3.5 397B A17B	66	—	88	53	65	57
140	Writer: Palmyra X5	66	—	74	71	—	52
150	Anthropic: Claude Sonnet 4.5	65	70	81	70	—	37
150	AionLabs: Aion 3.5	65	—	98	53	—	43
150	Pareto 26.9: Pareto 26.9	65	—	98	53	—	43
150	Perceptron: Perceptron Mk1.5	65	—	98	28	—	70
150	Qwen: Qwen3.6 35B A3B	65	—	90	53	39	76
150	Google: Gemma 4 31B	65	—	90	53	37	79
150	Anthropic: Claude Sonnet 4.6	65	—	88	70	—	37
150	Qwen: Qwen3.5-122B-A10B	65	—	88	53	52	65
150	Z.ai: GLM 4.7	65	—	85	50	—	61
150	MoonshotAI: Kimi K2 Thinking	65	—	83	53	—	60
150	ByteDance Seed: Seed 1.6	65	—	77	53	—	66
150	StepFun: Step 3.5 Flash	65	—	—	53	57	86
162	Qwen: Qwen3 Next 80B A3B Instruct	64	81	—	53	48	74
162	DeepSeek: DeepSeek V3.1 Terminus	64	73	—	47	—	71
162	MoonshotAI: Kimi K2 0905	64	68	—	53	75	60
162	MiniMax: MiniMax M1	64	60	—	70	—	61
162	OpenAI: gpt-oss-120b	64	60	—	44	52	100
162	IBM: Granite 4.2 8B	64	—	97	44	23	90
162	Sakana: Fugu Ultra	64	—	93	70	—	29
162	Anthropic: Claude Opus 4.8	64	—	92	70	—	30
162	Anthropic: Claude Opus 4.7	64	—	91	70	—	30
162	Z.ai: GLM 5V Turbo	64	—	90	50	—	52
162	Z.ai: GLM 5.1	64	—	90	50	—	51
162	Z.ai: GLM 5 Turbo	64	—	89	50	—	52
162	Qwen: Qwen3.5-35B-A3B	64	—	88	53	39	74
162	Z.ai: GLM 5	64	—	88	50	—	55
162	AionLabs: Aion-2.0	64	—	88	44	—	61
162	MiniMax: MiniMax M2-her	64	—	87	36	—	69
178	OpenAI: GPT-5.5 Pro	63	84	91	71	—	5
178	Qwen: Qwen3 30B A3B Instruct 2507	63	77	—	53	37	86
178	Qwen: Qwen3 VL 235B A22B Instruct	63	73	—	53	59	67
178	DeepSeek: R1 0528	63	73	—	47	71	62
178	Z.ai: GLM 4.5 Air	63	68	—	44	—	76
178	OpenAI: GPT-6 Astra Pro	63	—	97	71	—	21
178	Perceptron: Perceptron Mk1	63	—	92	27	—	70
178	Mistral: Mistral Medium 3.5	63	—	91	53	—	46
178	Mistral: Mistral Large 3 2512	63	—	—	53	71	64
187	Mistral: Mistral Medium 3.1	62	77	—	44	—	64
187	Google: Gemini 2.5 Pro	62	70	—	71	—	44
187	Google: Gemini 2.5 Pro Preview 06-05	62	70	—	71	—	44
187	MoonshotAI: Kimi K2 0711	62	68	—	44	75	61
187	OpenAI: GPT-5	62	64	79	59	—	44
187	DeepSeek: DeepSeek V3 0324	62	61	—	47	71	70
187	OpenAI: GPT-4.1	62	60	73	71	—	44
187	Meta: Muse Glimmer 30B	62	—	96	44	37	69
187	Qwen: Qwen3.5-9B	62	—	89	53	24	82
187	OpenAI: GPT-5.3-Codex	62	—	88	59	—	40
187	Anthropic: Claude Opus 4.6	62	—	87	70	—	30
187	OpenAI: GPT-5.2-Codex	62	—	86	59	—	40
187	OpenAI: GPT-5.1-Codex-Max	62	—	84	59	—	44
187	OpenAI: GPT-5.1	62	—	83	59	—	44
187	OpenAI: GPT-5.1-Codex	62	—	83	59	—	44
187	Qwen: Qwen3 Coder Next	62	—	70	53	48	77
203	Qwen: Qwen3 Coder 30B A3B Instruct	61	77	—	53	37	78
203	Qwen: Qwen3 32B	61	73	—	44	38	87
203	Anthropic: Claude Fable 5	61	—	93	70	—	21
203	OpenAI: GPT-5.2	61	—	85	59	—	40
203	Anthropic: Claude Haiku 4.5	61	—	81	50	—	51
203	Mistral: Mistral Small 4	61	—	—	53	52	78
209	Qwen: Qwen3 235B A22B Thinking 2507	60	77	—	44	59	58
209	Qwen: Qwen3 VL 30B A3B Instruct	60	73	—	53	37	78
209	Mistral: Mistral Medium 3	60	73	—	44	—	64
209	DeepSeek: DeepSeek V3	60	61	—	47	—	71
209	AionLabs: Aion-3.0	60	—	94	44	—	43
209	OpenAI: GPT Chat Latest	60	—	91	59	—	29
209	Qwen: Qwen3.6 27B	60	—	91	53	36	60
209	Qwen: Qwen3.5-27B	60	—	88	53	36	63
209	NVIDIA: Nemotron 3 Nano 30B A3B	60	—	—	53	37	90
209	MiniMax: MiniMax M2	60	—	—	50	59	71

40–59 overall score

Ranked by overall score, 40–59
Rank	Model	Overall	Freshness	Recency	Context	Parameters	Price
219	Qwen: Qwen3 VL 30B A3B Thinking	59	73	—	53	37	71
219	Amazon: Nova Pro 1.0	59	66	—	55	—	56
219	OpenAI: gpt-oss-20b	59	60	—	44	33	99
219	Cohere: Command R (08-2024)	59	56	—	44	—	78
219	LiquidAI: LFM2.5-24B-A2B	59	—	99	44	10	81
224	Z.ai: GLM 4.5	58	68	—	44	—	61
224	Meta: Llama 4 Maverick	58	63	—	62	31	76
224	OpenAI: GPT-4o-mini	58	48	61	44	—	78
227	Reka Flash 3: Reka Flash 3	57	70	—	36	33	88
227	Nous: Hermes 4 405B	57	63	—	44	65	55
227	Google: Gemma 3 27B	57	63	—	44	36	86
227	Google: Gemma 3 12B	57	63	—	44	27	95
227	OpenAI: o3	57	60	73	50	—	44
227	Microsoft: Phi 4	57	60	—	18	—	93
227	TheDrummer: UnslopNemo 12B	57	57	—	71	27	73
227	Mistral: Mistral Nemo	57	57	—	44	27	100
227	Meta: Llama 3.3 70B Instruct	57	51	—	44	46	85
227	Mistral: Mistral Small 3.2 24B	57	48	—	53	35	90
237	Tencent: Hunyuan A13B Instruct	56	73	—	44	28	79
237	ByteDance: UI-TARS 7B	56	70	—	44	21	88
237	LiquidAI: LFM2.5-2.6B	56	—	96	44	10	73
237	OpenAI: GPT-5.2 Chat	56	—	85	44	—	40
237	NVIDIA: Nemotron 3 Super	56	—	37	71	52	63
237	Mistral: Devstral 2 2512	56	—	—	53	52	64
243	Qwen: Qwen3 30B A3B	55	73	—	30	37	81
243	Qwen: Qwen3 14B	55	73	—	30	29	86
243	Meta: Llama Guard 4 12B	55	63	—	47	27	83
243	Google: Gemma 3 4B	55	63	—	44	15	97
243	OpenAI: GPT-5.4 Pro	55	—	89	71	—	5
243	Anthropic: Claude Opus 4.5	55	—	84	50	—	30
243	Mistral: Ministral 3 14B 2512	55	—	—	53	29	82
243	OpenAI: gpt-oss-safeguard-20b	55	—	—	44	33	87
251	Anthropic: Claude Opus 4.1	54	70	79	50	—	16
251	Z.ai: GLM 4.5V	54	68	—	36	51	62
251	OpenAI: GPT-5 Pro	54	64	82	59	—	12
251	Meta: Llama 3.1 8B Instruct	54	51	—	44	23	98
251	Mistral: Ministral 3 8B 2512	54	—	—	53	23	86
256	Qwen: Qwen2.5 VL 72B Instruct	53	60	—	44	46	63
256	TheDrummer: Cydonia 24B V4.1	53	57	—	44	35	74
256	Moondream: Moondream 3 Preview	53	—	99	26	24	63
256	Moondream: Moondream 3.1	53	—	99	26	24	63
260	Mistral Large: Mistral Large	52	67	—	44	—	46
260	Cohere: Command A	52	63	—	53	51	41
260	Qwen2.5 72B Instruct: Qwen2.5 72B Instruct	52	60	—	27	46	74
260	Venice: Uncensored	52	57	—	44	35	73
260	Meta: Llama 3.1 70B Instruct	52	51	—	44	46	65
260	Mistral: Mistral Small 3	52	48	—	27	35	98
266	Mistral: Saba	51	64	—	27	35	76
266	Sao10K: Llama 3.1 Euryale 70B v2.2	51	51	—	44	46	63
268	OpenAI: o3 Pro	50	60	76	50	—	14
268	OpenAI: GPT-4o	50	48	67	44	—	41
268	Mistral: Mistral Small 3.1 24B	50	48	—	44	35	73
268	OpenAI: GPT-5.2 Pro	50	—	85	59	—	7
268	Qwen: Qwen3 VL 8B Instruct	50	—	—	53	23	73
273	Qwen: Qwen2.5 7B Instruct	49	60	—	27	21	88
273	Mistral Large 2407: Mistral Large 2407	49	56	—	44	—	46
273	Meta: Llama 3.2 3B Instruct	49	51	—	44	12	88
273	Mistral: Ministral 3 3B 2512	49	—	—	44	12	91
273	Mistral: Voxtral Small 24B 2507	49	—	—	27	35	86
278	TheDrummer: Skyfall 36B V2	48	60	—	27	39	67
278	Mistral: Mixtral 8x22B Instruct	48	53	—	36	56	46
280	Qwen2.5 Coder 32B Instruct: Qwen2.5 Coder 32B Instruct	47	60	—	27	38	64
280	Cohere: Command R+ (08-2024)	47	56	—	44	—	41
282	Sao10K: Llama 3 8B Lunaris	46	51	—	9	23	100
283	Magnum v4 72B: Magnum v4 72B	45	60	—	27	46	46
283	Meta: Llama 3.2 1B Instruct	45	51	—	34	0	95
285	Google: Gemma 2 27B	43	60	—	9	36	66
285	MythoMax 13B: MythoMax 13B	43	43	—	9	28	93
287	AionLabs: Aion-RP 1.0 (8B)	41	51	—	27	23	61
287	Mancer: Weaver (alpha)	41	43	—	9	—	70

🤖 Generate website with AI 👆

Join the beta waitlist and get 50% off your first 3 months.

Join waitlist

20–39 overall score

Ranked by overall score, 20–39
Rank	Model	Overall	Freshness	Recency	Context	Parameters	Price
289	ReMM SLERP 13B: ReMM SLERP 13B	37	43	—	5	28	72
290	OpenAI: GPT-3.5 Turbo 16k	25	13	—	18	—	45
291	OpenAI: GPT-3.5 Turbo (older v0613)	24	13	—	0	—	58

0–19 overall score

Ranked by overall score, 0–19
	Rank	Model	Overall	Freshness	Recency	Context	Parameters	Price
	292	OpenAI: GPT-4	12	13	—	9	—	13

🤖 Generate website with AI 👆

Join the beta waitlist and get 50% off your first 3 months.

Join waitlist

How the overall score is calculated

The overall score is the average of the component scores that a model has. Each component is scaled from 0 to 100, and each one has its own ranking on this site.

Freshness and recency rise as the date moves toward today. Context and parameter count rise with size. The price score rises as the blended token price falls.

Freshness is ranked in AI Knowledge Cutoff. A more recent knowledge cutoff scores higher.
Recency is ranked in AI Release Date. A newer release scores higher.
Context is ranked in AI Context Window. A larger context window scores higher.
Parameters is ranked in AI Parameter Count. A larger published parameter count scores higher.
Price is ranked in AI Token Price. A cheaper blended token price scores higher.

A model needs at least three published scores. A missing score is left out of the average, not counted as zero. Ties are broken by the component scores in order.

This is not a quality benchmark. It does not measure reasoning, coding, or chat preference. It summarizes token price, context length, parameter count, release recency, and knowledge freshness.

Frequently asked questions

How are these AI models ranked?

Each ranked AI model gets one number from 0 to 100 that averages the published scores for knowledge freshness, release recency, context window, parameter count, and token price. A higher score ranks higher. A model is included when at least three of the five scores exist.

How is the overall score calculated?

Add the published component scores and divide by how many there are, then round to the nearest integer. A model needs at least three scores. Each component is already scaled from 0 to 100. Freshness and recency rise as the date moves toward today. Context and parameter count rise with size. The price score rises as the blended token price falls.

Which AI model has the highest overall score?

GLM 5.3 Flash has the highest overall score in this database, 89 out of 100.

Does a higher score mean a better chatbot?

No. The overall score does not measure reasoning, coding, or preference in chat. It summarizes price, context length, parameter count, release recency, and knowledge freshness.

Ranked AI ModelsShare

Ranked by overall score

80–100 overall score

🤖 Generate website with AI 👆

60–79 overall score

40–59 overall score

🤖 Generate website with AI 👆

20–39 overall score

0–19 overall score

🤖 Generate website with AI 👆

How the overall score is calculated

Frequently asked questions

How are these AI models ranked?

How is the overall score calculated?

Which AI model has the highest overall score?

Does a higher score mean a better chatbot?

Ranked AI Models