Cortex Lab

LLM performance leaderboard

Price, output speed and quality across every major model, compared side by side. This is the table we open before deciding what a workflow should run on. The data is published by Artificial Analysis — we host it here for convenience and credit it in full.

			Quality index			Cost $/1M tokens			Speed			Benchmarks
#	Model
1	Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback) Anthropic	Sep 2026	57.6	—	—	$8.00	$4.00	$20.00	96	477.23	477.23	—	61.4%	66.9%	84.7%	—	—	—	—	—	—	—	—
2	Claude Opus 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback) Anthropic	Sep 2026	56.0	—	—	$8.00	$4.00	$20.00	81	53.97	53.97	—	57.5%	65.0%	84.7%	—	—	—	—	—	—	—	—
3	Claude Sonnet 5.5 (Adaptive Reasoning, Max Effort, Default Fallback) Anthropic	Sep 2026	56.0	—	—	$4.00	$2.00	$10.00	145	308.29	308.29	—	55.0%	61.0%	82.7%	—	—	—	—	—	—	—	—
4	Claude Opus 5.5 (Adaptive Reasoning, High Effort, Default Fallback) Anthropic	Sep 2026	53.6	—	—	$8.00	$4.00	$20.00	74	15.65	15.65	—	55.6%	60.4%	82.7%	—	—	—	—	—	—	—	—
5	Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) Anthropic	Sep 2026	53.4	81.6	—	$20.00	$10.00	$50.00	69	114.39	114.39	93.7%	59.1%	63.1%	85.3%	—	—	—	—	—	—	91.4%	47.2%
6	Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback) Anthropic	Sep 2026	53.2	80.7	—	$20.00	$10.00	$50.00	60	34.30	34.30	93.4%	58.7%	60.9%	83.0%	—	—	—	—	—	—	91.0%	45.8%
7	GPT-6 Astra (max) OpenAI	Sep 2026	52.7	76.9	—	$20.00	$10.00	$50.00	59	177.90	177.90	96.1%	54.7%	56.5%	80.7%	—	—	—	—	—	—	88.4%	41.4%
8	GPT-6 Astra (xhigh) OpenAI	Sep 2026	52.4	75.9	—	$20.00	$10.00	$50.00	54	74.01	74.01	96.3%	54.6%	55.7%	80.0%	—	—	—	—	—	—	89.1%	43.1%
9	Claude Sonnet 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback) Anthropic	Sep 2026	51.9	—	—	$4.00	$2.00	$10.00	110	15.04	15.04	—	50.0%	57.3%	79.7%	—	—	—	—	—	—	—	—
10	GPT-6.1 Sol (max) OpenAI	Sep 2026	51.8	—	—	$4.00	$2.00	$10.00	87	118.51	118.51	—	52.9%	54.2%	83.0%	—	—	—	—	—	—	—	—
11	Claude Opus 5.5 (Adaptive Reasoning, Medium Effort, Default Fallback) Anthropic	Sep 2026	51.2	—	—	$8.00	$4.00	$20.00	74	12.75	12.75	—	54.7%	59.3%	84.3%	—	—	—	—	—	—	—	—
12	Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback) Anthropic	Sep 2026	51.2	79.1	—	$20.00	$10.00	$50.00	51	6.23	6.23	90.6%	55.9%	58.7%	83.7%	—	—	—	—	—	—	89.9%	43.1%
13	GPT-6.1 Sol (xhigh) OpenAI	Sep 2026	51.0	—	—	$4.00	$2.00	$10.00	89	37.54	37.54	—	52.6%	55.7%	79.7%	—	—	—	—	—	—	—	—
14	GPT-6 Astra (high) OpenAI	Sep 2026	50.9	77.1	—	$20.00	$10.00	$50.00	52	29.69	29.69	94.9%	53.1%	55.4%	80.0%	—	—	—	—	—	—	89.9%	40.0%
15	Claude Opus 5 (Adaptive Reasoning, Max Effort) Anthropic	Jul 2026	50.8	78.0	—	$10.00	$5.00	$25.00	—	—	—	93.2%	54.9%	56.4%	79.3%	—	—	—	—	—	—	89.1%	42.1%
16	GPT-6.1 Sol (high) OpenAI	Sep 2026	50.2	—	—	$4.00	$2.00	$10.00	78	19.18	19.18	—	51.4%	55.8%	82.3%	—	—	—	—	—	—	—	—
17	Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) Anthropic	Jul 2026	49.7	77.0	—	$10.00	$5.00	$25.00	—	—	—	93.7%	54.4%	55.7%	80.3%	—	—	—	—	—	—	88.0%	43.3%
18	Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) Anthropic	Jun 2026	49.6	76.5	—	$20.00	$10.00	$50.00	—	—	—	92.6%	55.5%	61.0%	82.3%	63.5%	98.5%	62.9%	—	—	—	84.6%	38.1%
19	GPT-6 Astra (medium) OpenAI	Sep 2026	49.6	76.7	—	$20.00	$10.00	$50.00	52	3.58	3.58	93.9%	52.7%	54.2%	79.7%	—	—	—	—	—	—	89.5%	35.5%
20	Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) Anthropic	Sep 2026	48.9	77.1	—	$20.00	$10.00	$50.00	50	4.80	4.80	88.6%	53.8%	56.4%	84.7%	—	—	—	—	—	—	88.0%	41.0%
21	Claude Opus 5 (Adaptive Reasoning, High Effort) Anthropic	Jul 2026	48.1	76.5	—	$10.00	$5.00	$25.00	—	—	—	93.7%	52.8%	55.4%	79.0%	—	—	—	—	—	—	87.6%	44.7%
22	Muse Spark 1.3 (max) Meta	Sep 2026	48.1	75.8	—	$2.00	$1.25	$4.25	226	20.92	29.78	93.5%	48.7%	58.8%	83.0%	—	—	—	—	—	—	84.3%	50.5%
23	GPT-6.1 Sol (medium) OpenAI	Sep 2026	47.8	—	—	$4.00	$2.00	$10.00	74	4.13	4.13	—	49.9%	53.2%	83.3%	—	—	—	—	—	—	—	—
24	GPT-6 Sol (max) OpenAI	Sep 2026	47.5	—	—	$4.00	$2.00	$10.00	83	108.52	108.52	—	47.9%	57.6%	83.7%	—	—	—	—	—	—	—	—
25	GPT-5.6 Sol (max) OpenAI	Jul 2026	47.0	77.4	—	$8.00	$4.00	$20.00	—	—	—	94.1%	49.5%	57.1%	84.0%	72.7%	85.1%	65.9%	—	—	—	88.0%	44.3%
26	Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback) Anthropic	Sep 2026	46.8	75.2	—	$20.00	$10.00	$50.00	51	2.95	2.95	88.1%	48.9%	56.7%	82.3%	—	—	—	—	—	—	85.0%	39.0%
27	Claude Sonnet 5.5 (Adaptive Reasoning, High Effort, Default Fallback) Anthropic	Sep 2026	46.7	—	—	$4.00	$2.00	$10.00	106	7.65	7.65	—	45.8%	53.7%	78.0%	—	—	—	—	—	—	—	—
28	Grok 4.7 (xhigh) SpaceXAI	Sep 2026	46.4	—	—	$3.00	$2.00	$6.00	83	19.84	19.84	—	43.1%	57.4%	76.7%	—	—	—	—	—	—	—	—
29	Grok 4.7 (high) SpaceXAI	Sep 2026	46.3	—	—	$3.00	$2.00	$6.00	79	20.37	20.37	—	42.3%	57.8%	77.0%	—	—	—	—	—	—	—	—
30	MiMo-V2.6-Pro Xiaomi	Sep 2026	46.3	—	—	$0.54	$0.44	$0.87	43	3.11	49.73	—	49.4%	60.9%	86.3%	—	—	—	—	—	—	—	—
31	GPT-6 Astra (low) OpenAI	Sep 2026	45.8	75.7	—	$20.00	$10.00	$50.00	50	1.71	1.71	93.1%	49.2%	54.1%	80.0%	—	—	—	—	—	—	88.0%	32.0%
32	Qwen3.8 Max (0902) Alibaba	Sep 2026	45.4	76.2	—	$3.00	$2.00	$6.00	38	1.91	54.12	92.8%	43.1%	52.1%	80.3%	—	—	—	—	—	—	88.8%	47.8%
33	Muse Spark 1.3 (xhigh) Meta	Sep 2026	45.1	76.5	—	$2.00	$1.25	$4.25	153	41.04	54.12	94.1%	47.5%	59.7%	83.0%	—	—	—	—	—	—	85.4%	47.2%
34	Claude Opus 5 (Adaptive Reasoning, Medium Effort) Anthropic	Jul 2026	44.8	74.3	—	$10.00	$5.00	$25.00	—	—	—	91.9%	51.3%	51.5%	82.0%	—	—	—	—	—	—	86.1%	38.6%
35	GLM-5.3 (max) Z AI	Aug 2026	44.8	74.8	—	$2.15	$1.40	$4.40	75	2.70	29.29	91.7%	42.3%	59.0%	79.7%	—	—	—	—	—	—	83.9%	50.3%
36	Grok 4.6 (high) SpaceXAI	Aug 2026	44.3	76.8	—	$3.00	$2.00	$6.00	63	22.79	22.79	94.9%	42.9%	56.5%	80.3%	—	—	—	—	—	—	88.4%	50.7%
37	Grok 4.6 (xhigh) SpaceXAI	Aug 2026	44.2	75.9	—	$3.00	$2.00	$6.00	61	25.42	25.42	93.5%	44.1%	53.0%	81.0%	—	—	—	—	—	—	88.0%	43.3%
38	GPT-6 Sol (xhigh) OpenAI	Sep 2026	44.1	—	—	$4.00	$2.00	$10.00	74	33.58	33.58	—	46.3%	55.1%	81.3%	—	—	—	—	—	—	—	—
39	GPT-5.6 Sol (xhigh) OpenAI	Jul 2026	44.0	78.3	—	$8.00	$4.00	$20.00	—	—	—	93.1%	47.3%	57.1%	82.3%	71.0%	84.8%	61.4%	—	—	—	89.5%	38.1%
40	Step 5 Preview StepFun	Sep 2026	43.7	—	—	$1.43	$1.00	$2.70	86	1.97	25.12	—	46.5%	58.9%	88.3%	—	—	—	—	—	—	—	—
41	Kimi K3 (max) Kimi	Jul 2026	43.6	76.2	—	$6.00	$3.00	$15.00	42	6.38	54.41	93.5%	46.9%	59.5%	88.7%	—	—	—	—	—	—	85.0%	46.0%
42	Grok 4.6 (medium) SpaceXAI	Aug 2026	42.8	74.4	—	$3.00	$2.00	$6.00	61	21.09	21.09	93.5%	42.1%	55.9%	81.0%	—	—	—	—	—	—	84.3%	44.3%
43	GPT-6 Sol (high) OpenAI	Sep 2026	42.8	—	—	$4.00	$2.00	$10.00	73	16.07	16.07	—	44.1%	54.9%	83.7%	—	—	—	—	—	—	—	—
44	GPT-5.6 Sol (high) OpenAI	Jul 2026	42.3	77.2	—	$8.00	$4.00	$20.00	—	—	—	92.8%	46.0%	57.8%	81.7%	69.2%	83.3%	62.1%	—	—	—	87.3%	36.7%
45	Claude Opus 5.5 (Adaptive Reasoning, Low Effort, Default Fallback) Anthropic	Sep 2026	42.3	—	—	$8.00	$4.00	$20.00	75	4.85	4.85	—	48.3%	58.6%	80.7%	—	—	—	—	—	—	—	—
46	GPT-6.1 Sol (low) OpenAI	Sep 2026	42.1	—	—	$4.00	$2.00	$10.00	81	1.23	1.23	—	47.4%	53.2%	84.0%	—	—	—	—	—	—	—	—
47	GPT-5.6 Terra (max) OpenAI	Jul 2026	42.1	76.7	—	$4.50	$2.00	$12.00	106	86.60	86.60	92.5%	42.9%	55.0%	83.0%	71.2%	86.3%	57.6%	—	—	—	88.0%	40.2%
48	GLM 5.3 Flash Z AI	Aug 2026	41.8	71.5	—	$0.24	$0.15	$0.50	47	3.16	45.77	91.2%	39.9%	51.6%	80.0%	—	—	—	—	—	—	84.3%	47.2%
49	Claude Opus 4.8 (Adaptive Reasoning, Max Effort) Anthropic	May 2026	41.8	74.3	—	$10.00	$5.00	$25.00	—	—	—	92.0%	48.7%	54.4%	77.7%	62.2%	94.4%	58.3%	—	—	—	84.6%	34.2%
50	Gemini 3.8 Flash (high) Google	Sep 2026	40.9	76.3	—	$1.50	$0.75	$3.75	230	12.04	12.04	95.3%	47.8%	56.6%	81.3%	—	—	—	—	—	—	87.6%	44.9%
51	Claude Sonnet 5.5 (Adaptive Reasoning, Medium Effort, Default Fallback) Anthropic	Sep 2026	40.7	—	—	$4.00	$2.00	$10.00	88	0.81	0.81	—	39.8%	52.9%	76.3%	—	—	—	—	—	—	—	—
52	Claude Opus 4.7 (Adaptive Reasoning, Max Effort) Anthropic	Apr 2026	40.7	73.6	—	$10.00	$5.00	$25.00	—	—	—	91.4%	42.3%	—	78.7%	58.6%	88.6%	51.5%	—	—	—	83.1%	34.6%
53	Qwen3.8 Max Alibaba	Aug 2026	40.2	71.8	—	$3.00	$2.00	$6.00	—	—	—	92.7%	43.0%	53.2%	78.3%	—	—	—	—	—	—	81.3%	51.3%
54	Qwen3.8 2.4T A95B Alibaba	Aug 2026	39.9	71.9	—	$3.00	$2.00	$6.00	38	2.02	54.16	93.5%	42.4%	54.1%	80.3%	—	—	—	—	—	—	82.0%	49.1%
55	Qwen3.8-Flash-Next Alibaba	Aug 2026	39.8	73.1	—	$0.23	$0.15	$0.47	53	1.53	39.31	92.3%	38.0%	50.6%	79.7%	—	—	—	—	—	—	86.1%	45.4%
56	GPT-6 Sol (medium) OpenAI	Sep 2026	39.8	—	—	$4.00	$2.00	$10.00	—	—	—	—	41.0%	53.8%	82.3%	—	—	—	—	—	—	—	—
57	Gemini 3.8 Flash (medium) Google	Sep 2026	39.8	74.1	—	$1.50	$0.75	$3.75	—	—	—	93.5%	42.1%	55.1%	84.0%	—	—	—	—	—	—	83.9%	45.8%
58	Gemini 3.7 Flash (medium) Google	Aug 2026	39.6	71.5	—	$1.50	$0.75	$3.75	—	—	—	92.1%	39.0%	59.8%	83.0%	—	—	—	—	—	—	78.3%	35.5%
59	Muse Spark 1.2 (xhigh) Meta	Aug 2026	39.6	72.2	—	$2.00	$1.25	$4.25	—	—	—	90.4%	45.5%	57.4%	79.0%	—	—	—	—	—	—	80.1%	34.8%
60	DeepSeek V4.1 Flash (Reasoning, Max Effort) DeepSeek	Sep 2026	39.5	—	—	$0.53	$0.30	$1.20	219	0.84	9.98	—	39.2%	51.9%	84.0%	—	—	—	—	—	—	—	—
61	Claude Opus 5 (Adaptive Reasoning, Low Effort) Anthropic	Jul 2026	39.4	66.9	—	$10.00	$5.00	$25.00	—	—	—	88.9%	43.4%	49.2%	81.3%	—	—	—	—	—	—	76.4%	30.3%
62	GPT-5.6 Sol (medium) OpenAI	Jul 2026	39.2	76.3	—	$8.00	$4.00	$20.00	—	—	—	92.6%	42.2%	57.4%	80.3%	69.6%	81.0%	62.9%	—	—	—	86.1%	36.5%
63	Gemini 3.7 Flash (high) Google	Aug 2026	39.1	76.1	—	$1.50	$0.75	$3.75	—	—	—	94.5%	47.9%	57.2%	81.7%	—	—	—	—	—	—	85.8%	32.8%
64	GPT-5.4 (xhigh) OpenAI	Mar 2026	39.0	71.1	—	$5.63	$2.50	$15.00	—	—	—	92.0%	43.7%	—	82.0%	73.9%	87.1%	57.6%	—	—	—	78.3%	39.6%
65	Grok 4.5 (high) SpaceXAI	Jul 2026	38.8	72.4	—	$3.00	$2.00	$6.00	—	—	—	93.1%	42.7%	55.0%	79.3%	—	—	—	—	—	—	81.6%	42.1%
66	GPT-5.5 (xhigh) OpenAI	Apr 2026	38.4	74.9	—	$11.25	$5.00	$30.00	—	—	—	93.5%	45.8%	55.8%	84.3%	75.9%	93.9%	60.6%	—	—	—	84.3%	39.0%
67	Claude Sonnet 5 (Adaptive Reasoning, Max Effort) Anthropic	Jun 2026	38.2	71.5	—	$4.00	$2.00	$10.00	87	123.05	123.05	91.1%	41.3%	54.3%	82.0%	—	—	—	—	—	—	80.5%	37.3%
68	GPT-5.6 Terra (xhigh) OpenAI	Jul 2026	38.0	70.6	—	$4.50	$2.00	$12.00	88	8.15	8.15	90.8%	41.9%	52.3%	79.0%	66.3%	80.4%	62.9%	—	—	—	80.1%	29.7%
69	MiMo-V2.6-Flash Xiaomi	Sep 2026	37.9	—	—	$0.18	$0.14	$0.28	47	4.07	46.20	—	35.1%	51.3%	74.3%	—	—	—	—	—	—	—	—
70	GPT-5.6 Luna (max) OpenAI	Jul 2026	37.3	71.4	—	$0.45	$0.20	$1.20	—	—	—	91.1%	39.5%	53.6%	83.7%	—	—	—	—	—	—	80.9%	31.1%
71	GPT-6 Luna (max) OpenAI	Sep 2026	37.3	—	—	$0.20	$0.10	$0.50	138	68.01	68.01	—	38.5%	54.6%	83.3%	—	—	—	—	—	—	—	—
72	GPT-5.5 (high) OpenAI	Apr 2026	37.0	71.6	—	$11.25	$5.00	$30.00	—	—	—	93.2%	45.0%	56.1%	84.3%	71.6%	93.0%	59.8%	—	—	—	79.4%	36.7%
73	Gemini 3.7 Flash (low) Google	Aug 2026	36.9	71.0	—	$1.50	$0.75	$3.75	—	—	—	90.1%	35.1%	55.7%	78.7%	—	—	—	—	—	—	79.8%	29.5%
74	DeepSeek V4 Pro 0813 (Reasoning, Max Effort) DeepSeek	Aug 2026	36.0	68.8	—	$1.98	$1.32	$3.96	87	1.12	24.07	92.8%	41.0%	51.0%	80.3%	—	—	—	—	—	—	78.7%	39.6%
75	Grok 4.6 (low) SpaceXAI	Aug 2026	35.1	66.3	—	$3.00	$2.00	$6.00	61	3.41	3.41	87.9%	27.6%	49.4%	80.7%	—	—	—	—	—	—	75.3%	38.1%
76	DeepSeek V4 Flash Vision (Reasoning, Max Effort) DeepSeek	Aug 2026	34.8	65.0	—	$0.66	$0.44	$1.32	222	0.77	9.78	91.3%	34.5%	49.7%	81.3%	—	—	—	—	—	—	74.2%	41.0%
77	GPT-5.6 Luna (xhigh) OpenAI	Jul 2026	34.6	68.6	—	$0.45	$0.20	$1.20	—	—	—	89.5%	37.0%	50.5%	81.7%	—	—	—	—	—	—	77.9%	28.7%
78	Kimi K3 (low) Kimi	Jul 2026	34.5	72.0	—	$6.00	$3.00	$15.00	40	3.62	54.21	84.2%	25.0%	52.7%	79.3%	—	—	—	—	—	—	82.4%	41.6%
79	Claude Sonnet 5 (Adaptive Reasoning, Xhigh Effort) Anthropic	Jun 2026	34.4	—	—	$4.00	$2.00	$10.00	76	15.77	15.77	—	39.0%	54.1%	76.7%	—	—	—	—	—	—	—	—
80	DeepSeek V4 Flash 0731 (Reasoning, Max Effort) DeepSeek	Jul 2026	34.3	69.1	—	$0.66	$0.44	$1.32	—	—	—	90.8%	38.6%	50.3%	79.7%	—	—	—	—	—	—	78.7%	39.4%
81	GLM-5.3 (low) Z AI	Aug 2026	34.3	—	—	$2.15	$1.40	$4.40	74	3.24	30.26	—	36.6%	42.0%	72.7%	—	—	—	—	—	—	—	—
82	GPT-5.6 Terra (high) OpenAI	Jul 2026	34.2	67.1	—	$4.50	$2.00	$12.00	90	2.18	2.18	89.6%	38.5%	52.4%	77.7%	64.4%	78.4%	57.6%	—	—	—	75.7%	28.7%
83	Gemini 3.6 Flash (high) Google	Jul 2026	34.0	69.2	—	$1.50	$0.75	$3.75	—	—	—	92.8%	40.8%	53.4%	80.0%	—	—	—	—	—	—	77.5%	29.9%
84	GPT-6 Sol (low) OpenAI	Sep 2026	33.9	—	—	$4.00	$2.00	$10.00	70	1.76	1.76	—	34.9%	50.2%	79.3%	—	—	—	—	—	—	—	—
85	GPT-6 Luna (xhigh) OpenAI	Sep 2026	33.9	—	—	$0.20	$0.10	$0.50	137	14.69	14.69	—	34.3%	51.7%	80.0%	—	—	—	—	—	—	—	—
86	GPT-5.5 (medium) OpenAI	Apr 2026	33.8	71.5	—	$11.25	$5.00	$30.00	—	—	—	92.6%	42.4%	54.5%	83.0%	71.0%	91.8%	57.6%	—	—	—	80.5%	29.9%
87	Muse Spark 1.1 (xhigh) Meta	Jul 2026	33.7	71.3	—	$2.00	$1.25	$4.25	—	—	—	89.8%	46.2%	58.8%	77.7%	—	—	—	—	—	—	77.9%	31.8%
88	GLM-5.2 (max) Z AI	Jun 2026	33.7	68.8	—	$2.15	$1.40	$4.40	—	—	—	89.5%	41.1%	51.2%	78.3%	73.3%	99.1%	50.8%	—	—	—	77.9%	34.6%
89	Qwen3.8 27B (xhigh) Alibaba	Aug 2026	33.7	68.1	—	$1.13	$0.50	$3.00	46	1.24	44.98	90.5%	33.9%	46.6%	82.0%	—	—	—	—	—	—	79.8%	48.0%
90	Gemini 3.5 Flash (medium) Google	May 2026	33.6	—	—	$3.38	$1.50	$9.00	—	—	—	92.1%	41.3%	—	74.3%	74.6%	95.6%	39.4%	—	—	—	—	—
91	Motif 3 Motif Technologies	Aug 2026	33.6	63.5	—	—	—	—	—	—	—	83.4%	37.0%	42.2%	76.0%	—	—	—	—	—	—	74.9%	35.3%
92	GPT-5.6 Sol (low) OpenAI	Jul 2026	33.5	69.7	—	$8.00	$4.00	$20.00	—	—	—	89.8%	39.4%	56.4%	78.0%	66.5%	76.0%	60.6%	—	—	—	76.8%	29.1%
93	Gemini 3.8 Flash (low) Google	Sep 2026	33.5	73.5	—	$1.50	$0.75	$3.75	—	—	—	92.0%	37.1%	55.0%	80.7%	—	—	—	—	—	—	83.1%	33.2%
94	Gemini 3.5 Flash (high) Google	May 2026	32.6	70.1	—	$3.38	$1.50	$9.00	—	—	—	92.2%	42.7%	53.9%	73.3%	76.3%	95.3%	40.9%	—	—	—	78.7%	32.2%
95	GPT-5.3 Codex (xhigh) OpenAI	Feb 2026	32.5	—	—	$4.81	$1.75	$14.00	105	37.74	37.74	91.5%	42.5%	—	83.3%	75.4%	86.0%	53.0%	—	—	—	—	—
96	Motif 3 (Beta) Motif Technologies	Jul 2026	32.3	62.0	—	—	—	—	—	—	—	86.9%	40.4%	—	74.7%	—	—	—	—	—	—	70.8%	—
97	GPT-6 Luna (high) OpenAI	Sep 2026	32.1	—	—	$0.20	$0.10	$0.50	140	6.92	6.92	—	32.9%	50.3%	79.3%	—	—	—	—	—	—	—	—
98	GPT-5.6 Luna (high) OpenAI	Jul 2026	32.1	63.3	—	$0.45	$0.20	$1.20	—	—	—	89.2%	33.4%	51.6%	80.3%	—	—	—	—	—	—	69.7%	25.2%
99	Claude Opus 4.6 (Adaptive Reasoning, Max Effort) Anthropic	Feb 2026	31.9	—	—	$10.00	$5.00	$25.00	—	—	—	89.6%	39.9%	—	78.0%	53.1%	92.1%	46.2%	—	—	—	—	—
100	Claude Sonnet 5 (Adaptive Reasoning, High Effort) Anthropic	Jun 2026	31.7	—	—	$4.00	$2.00	$10.00	61	2.77	2.77	—	35.7%	54.3%	76.7%	—	—	—	—	—	—	—	—
101	Muse Spark Meta	Apr 2026	31.3	58.6	—	—	—	—	—	—	—	88.4%	40.7%	—	78.0%	75.9%	91.5%	45.5%	—	—	—	62.2%	—
102	Claude Opus 4.7 (Non-reasoning, High Effort) Anthropic	Apr 2026	30.9	—	—	$10.00	$5.00	$25.00	—	—	—	88.5%	33.3%	—	75.7%	43.6%	74.0%	54.5%	—	—	—	—	—
103	GPT-5.5 (low) OpenAI	Apr 2026	30.7	60.9	—	$11.25	$5.00	$30.00	—	—	—	91.0%	32.7%	—	81.0%	64.4%	83.9%	52.3%	—	—	—	65.5%	24.9%
104	K2 Horizon 375B A23B Institute of Foundation Models	Sep 2026	30.5	61.5	—	—	—	—	—	—	—	87.3%	32.0%	42.9%	80.0%	—	—	—	—	—	—	71.9%	34.2%
105	DeepSeek V4 Pro 0424 (Reasoning, Max Effort) DeepSeek	Apr 2026	30.4	59.4	—	$0.54	$0.44	$0.87	—	—	—	88.8%	37.5%	50.8%	74.7%	76.5%	96.2%	46.2%	—	—	—	64.0%	30.1%
106	GPT-5.2 (xhigh) OpenAI	Dec 2025	30.4	—	99.0	$4.81	$1.75	$14.00	—	—	—	90.3%	37.7%	—	82.7%	75.4%	84.8%	47.0%	87.4%	88.9%	99.0%	—	—
107	DeepSeek V4 Pro 0424 (Reasoning, High Effort) DeepSeek	Apr 2026	30.1	58.7	—	$0.54	$0.44	$0.87	—	—	—	90.5%	35.2%	—	70.3%	71.3%	94.2%	41.7%	—	—	—	64.8%	26.2%
108	GPT-5.6 Terra (medium) OpenAI	Jul 2026	30.1	64.7	—	$4.50	$2.00	$12.00	95	1.79	1.79	87.2%	33.3%	50.5%	74.0%	62.2%	72.8%	—	—	—	—	72.3%	25.6%
109	Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) Anthropic	Feb 2026	30.1	63.0	—	$6.00	$3.00	$15.00	—	—	—	87.5%	33.6%	50.1%	80.0%	56.6%	75.7%	53.0%	—	—	—	71.2%	34.4%
110	Gemini 3.1 Pro Preview Google	Feb 2026	29.7	68.8	—	$4.50	$2.00	$12.00	131	21.08	21.08	94.1%	47.0%	58.7%	82.0%	77.1%	95.6%	53.8%	—	—	—	73.8%	21.4%
111	GPT-6 Luna (medium) OpenAI	Sep 2026	29.5	—	—	$0.20	$0.10	$0.50	—	—	—	—	28.3%	50.9%	78.3%	—	—	—	—	—	—	—	—
112	Qwen3.7 Max Alibaba	May 2026	29.5	66.0	—	$3.75	$2.50	$7.50	—	—	—	92.3%	40.5%	49.5%	79.0%	80.5%	94.7%	50.8%	—	—	—	74.5%	11.8%
113	MiniMax-M3 MiniMax	Jun 2026	29.2	58.6	—	$0.53	$0.30	$1.20	138	0.91	15.45	92.9%	39.0%	47.1%	83.0%	82.9%	88.9%	42.4%	—	—	—	65.2%	15.3%
114	Claude Opus 4.5 (Reasoning) Anthropic	Nov 2025	29.1	—	91.3	$10.00	$5.00	$25.00	—	—	—	86.6%	30.1%	—	77.3%	58.0%	89.5%	47.0%	89.5%	87.1%	91.3%	—	—
115	MiMo-V2-Pro Xiaomi	Mar 2026	28.6	—	—	—	—	—	—	—	—	87.0%	30.4%	—	68.3%	68.8%	95.0%	40.9%	—	—	—	—	—
116	GPT-5.2 Codex (xhigh) OpenAI	Dec 2025	28.5	—	—	$4.81	$1.75	$14.00	—	—	—	89.9%	35.7%	—	82.3%	77.6%	92.1%	37.1%	—	—	—	—	—
117	Qwen3.6 Max Preview Alibaba	Apr 2026	28.4	—	—	$2.93	$1.30	$7.80	—	—	—	88.8%	30.8%	—	80.7%	76.6%	95.9%	43.9%	—	—	—	—	—
118	GPT-5.6 Sol (Non-reasoning) OpenAI	Jul 2026	28.3	65.1	—	$8.00	$4.00	$20.00	—	—	—	79.0%	16.7%	47.7%	62.3%	—	—	—	—	—	—	74.2%	19.6%
119	Nex-N2-Pro (based on Qwen3.5-397B-A17B) Nex AGI	Jun 2026	28.2	59.1	—	—	—	—	—	—	—	89.2%	33.7%	43.3%	81.7%	66.2%	81.6%	—	—	—	—	67.8%	17.5%
120	Solar Pro 4 Upstage	Aug 2026	28.2	52.7	—	$0.53	$0.30	$1.20	101	1.35	21.11	89.1%	29.2%	44.6%	74.0%	—	—	—	—	—	—	57.3%	23.3%

Independent benchmarks & pricing of LLMs — we publish them here, we don’t produce them. For further analysis and methodology, see artificialanalysis.ai. Updated 30 Sep 2026.

How to read it

Three things that trip people up when they first use this table.

Price is per million tokens

Input and output are priced separately, and output is usually the expensive half. A workflow that reads a lot and writes a little costs far less than the headline number suggests.

Two different kinds of speed

Time to first token is what a person waiting on a chat reply feels. Tokens per second is what matters for a batch job running overnight. Optimising for the wrong one wastes money.

Quality scores are a shortlist, not a verdict

Aggregate indices are a good way to narrow twenty models to three. They are a poor way to choose between those three — for that, test them on your actual task.

Building a workflow? We pick the model for you

We cost and benchmark every automation before it ships — so it runs on the cheapest model that actually holds up on your task.

Browse ready-made workflows Back to the Lab