Skip to content
Model intelligence

236 AI models

Capability (Epoch Capabilities Index and individual benchmarks from Epoch AI), API list prices and context windows (OpenRouter), release dates and open-weight status. Benchmarks are shown individually - never merged into one invented score.
Leads on ECI
Claude Opus 5.5
ECI 167.35Epoch AI
Leads on GPQA Diamond
GPT-6 Astra
95.8%Epoch AI
Leads on SWE-bench Verified
Claude Opus 4.7
83.5%Epoch AI
Cheapest in ECI top 10
Claude Sonnet 5.5
$2.00/M inputOpenRouter
Longest context
Grok 4.20 Multi-Agent
2M tokensOpenRouter
Best open-weight on ECI
Kimi K3
ECI 157.61Epoch AI
Epoch Capabilities Index by release date

Capability frontier

Source: Epoch AI (CC BY 4.0). ECI is Epoch's item-response model over many benchmarks; each model page shows its 90% confidence interval.

Blended $/1M tokens (3:1 input:output) vs ECI

Price vs capability

Only models with both an ECI score and an OpenRouter price. Normalisation: 75% input + 25% output price. Click a point to open the model.

#ModelECI ↓ContextIn $/MReleased
1Claude Opus 5.5
Anthropic· reasoning
167.31M$4.0022 Sep 2026
2GPT-6 Astra
OpenAI· reasoning
166.51.05M$10.03 Sep 2026
3Claude Sonnet 5.5
Anthropic· reasoning
165.21M$2.0028 Sep 2026
4Claude Fable 5.1
Anthropic· reasoning
164.81M$10.01 Sep 2026
5Claude Opus 5
Anthropic· reasoning
162.91M$5.0024 Jul 2026
6GPT-5.5 Pro
OpenAI· reasoning
162.41.05M$30.023 Apr 2026
7Claude Fable 5
Anthropic· reasoning
162.21M$10.09 Jun 2026
8GPT-5.6 Sol
OpenAI· reasoning
161.81.05M$2.009 Jul 2026
9GPT-5.6 Terra
OpenAI· reasoning
159.81.05M$2.009 Jul 2026
10GPT-5.5
OpenAI· reasoning
159.21.05M$5.0023 Apr 2026
11GPT-5.4 Pro
OpenAI· reasoning
159.11.05M$30.05 Mar 2026
12Claude Opus 4.8
Anthropic· reasoning
158.31M$5.0028 May 2026
13Gemini 3.7 Flash
Google· reasoning
157.41.05M$0.7513 Aug 2026
14Gemini 3.8 Flash
Google· reasoning
156.91.05M$0.752 Sep 2026
15GPT-5.4
OpenAI· reasoning
156.91.05M$2.505 Mar 2026
16Muse Spark 1.3
Meta· reasoning
156.91.05M$1.252 Sep 2026
17GPT-5.3-Codex
OpenAI· reasoning
156.8400K$1.755 Feb 2026
18Grok 4.6
xAI· reasoning
156.6500K$2.0012 Aug 2026
19Qwen 3.8 Max156.6——2 Aug 2026
20GPT-5.6 Luna
OpenAI· reasoning
156.41.05M$0.209 Jul 2026
21Claude Opus 4.7
Anthropic· reasoning
156.31M$5.0016 Apr 2026
22Claude Sonnet 5
Anthropic· reasoning
156.31M$2.0030 Jun 2026
23GPT-5.2 Pro
OpenAI· reasoning
155.4400K$21.011 Dec 2025
24Claude Opus 4.6
Anthropic· reasoning
155.31M$5.005 Feb 2026
25Qwen3.8 Max (0902)
Alibaba (Qwen)· reasoning
155.21M$2.001 Sep 2026
26Muse Spark 1.2
Meta· reasoning
155.01.05M$1.255 Aug 2026
27Gemini 3.1 Pro154.8——19 Feb 2026
28Gemini 3.5 Flash
Google· reasoning
154.51.05M$1.5019 May 2026
29Gemini 3.6 Flash
Google· reasoning
154.31.05M$0.7521 Jul 2026
30Muse Spark 1.1
Meta· reasoning
154.31.05M$1.259 Jul 2026
31Grok 4.5
xAI· reasoning
154.0500K$2.008 Jul 2026
32Qwen3.7 Max
Alibaba (Qwen)· reasoning
153.71M$1.4819 May 2026
33GPT-5.2
OpenAI· reasoning
153.5400K$1.7511 Dec 2025
34Gemini 3 Pro152.9——18 Nov 2025
35Claude Sonnet 4.6
Anthropic· reasoning
152.31M$3.0017 Feb 2026
36Muse Spark152.1——8 Apr 2026
37Grok 4.20
xAI· reasoning
152.02M$1.2517 Feb 2026
38Gemini 3 Flash151.8——17 Dec 2025
39GPT-5 Pro
OpenAI· reasoning
150.3400K$15.07 Oct 2025
40Claude Opus 4.5
Anthropic· reasoning
150.1200K$5.0024 Nov 2025
41GPT-5
OpenAI· reasoning
150.0400K$1.257 Aug 2025
42GPT-5.1
OpenAI· reasoning
149.7400K$1.2513 Nov 2025
43Qwen 3.6 Max (Preview)149.2——20 Apr 2026
44Grok 4.3 Beta149.2——17 Apr 2026
45GPT-5.4 Mini
OpenAI· reasoning
148.8400K$0.7517 Mar 2026
46Qwen 3.6 Plus147.6——31 Mar 2026
47o3 Pro
OpenAI· reasoning
147.4200K$20.010 Jun 2025
48Qwen3.7 Plus
Alibaba (Qwen)· reasoning
147.41M$0.322 Jun 2026
49o3
OpenAI· reasoning
146.9200K$2.0016 Apr 2025
50Claude Sonnet 4.5
Anthropic· reasoning
146.81M$3.0029 Sep 2025
51Qwen 3.5 Plus (hosted 397B-A17B)146.8——16 Feb 2026
52Grok 4146.4——9 Jul 2025
53GPT-5.4 Nano
OpenAI· reasoning
145.8400K$0.2017 Mar 2026
54o4 Mini
OpenAI· reasoning
145.6200K$1.1016 Apr 2025
55GPT-5 Mini
OpenAI· reasoning
145.5400K$0.257 Aug 2025
56Gemini 2.5 Pro (Jun 2025)145.3——5 Jun 2025
57Gemini 3.5 Flash Lite
Google· reasoning
145.11.05M$0.3021 Jul 2026
58Qwen3.7 Flash
Alibaba (Qwen)· reasoning
144.61M$0.03027 Jul 2026
59Gemini 3.1 Flash Lite
Google· reasoning
144.41.05M$0.253 Mar 2026
60Grok 4 Fast144.2——19 Sep 2025
61Gemini 2.5 Pro (Mar 2025)144.2——31 Mar 2025
62Claude Opus 4.1
Anthropic· reasoning
144.1200K$15.05 Aug 2025
63Qwen 3.5 Flash (hosted 35B-A3B)144.0——25 Feb 2026
64Qwen 3.6 Flash143.3——27 Apr 2026
65Gemini 2.5 Flash (Sep 2025)143.0——25 Sep 2025
66Claude Opus 4142.7——22 May 2025
67GPT-5.5 Instant142.5——5 May 2026
68Gemini 2.5 Pro (May 2025)142.5——6 May 2025
69Claude Haiku 4.5
Anthropic· reasoning
142.4200K$1.0015 Oct 2025
70Qwen3 Max
Alibaba (Qwen)· reasoning
142.4262K$0.7824 Sep 2025
71o1
OpenAI· reasoning
141.9200K$15.017 Dec 2024
72Claude Sonnet 4
Anthropic· reasoning
141.7200K$3.0022 May 2025
73Gemini 2.5 Flash (May 2025)141.5——20 May 2025
74Claude 3.7 Sonnet141.2——24 Feb 2025
75Gemini 2.5 Flash (Jun 2025)140.8——17 Jun 2025
76Grok-3 mini140.3——24 Jun 2025
77o3 Mini
OpenAI· reasoning
140.3200K$1.1031 Jan 2025
78Gemini 2.5 Flash (Apr 2025)140.0——17 Apr 2025
79GPT-5 Nano
OpenAI· reasoning
139.4400K$0.0507 Aug 2025
80Grok 3138.3——9 Apr 2025
81GPT-4.1136.81.05M$2.0014 Apr 2025
82GPT-4.5136.7——27 Feb 2025
83o1-mini135.8——12 Sep 2024
84Gemini 2.0 Flash Thinking (Jan 2025)135.4——21 Jan 2025
85Gemini 2.0 Pro135.1——5 Feb 2025
86GPT-4.1 Mini135.01.05M$0.4014 Apr 2025
87o1-preview134.8——12 Sep 2024
88Gemini 2.0 Flash (Feb 2025)134.7——5 Feb 2025
89Gemini 2.0 Flash (Dec 2024)134.7——11 Dec 2024
90Mistral Medium 3134.1131K$0.407 May 2025
91Gemini 2.5 Flash-Lite (Jun 2025)133.9——17 Jun 2025
92Claude 3.5 Sonnet (October 2024)133.5——22 Oct 2024
93Qwen2.5-Max132.5——25 Jan 2025
94Gemini 1.5 Pro (Sept 2024)131.7——24 Sep 2024
95Grok-2 (Dec 2024)130.5——12 Dec 2024
96Claude 3.5 Sonnet130.0——20 Jun 2024
97GPT-4.1 Nano129.61.05M$0.1014 Apr 2025
98Gemini 1.5 Flash (Sep 2024)129.4——24 Sep 2024
99GPT-4o (May 2024)129.0——13 May 2024
100GPT-4o (Nov 2024)128.8——20 Nov 2024
Showing 100 of 236. Benchmark cells are % (best reported setting). 3.9K benchmark results in total.Next →Export CSV