Skip to content

Best AI Models for Agents

Building AI agents? These 394 models support function calling and are ranked by their agentic capability coverage - tool use, reasoning, JSON output, streaming, vision, and web search.

394
Agent-Ready
311
+ Reasoning
354
+ JSON Mode
159
+ Web Search
23
Free

Full-Stack Agent Models

5+ agentic capabilities
#ModelScoreCaps
1Claude Fable 5.1Anthropic966
2Claude Fable 5.1 (batch)Anthropic966
3Claude Fable 5Anthropic966
4Claude Fable 5 (batch)Anthropic966
5Claude Opus 5.5Anthropic956
6Claude Opus 5Anthropic956
7Claude Opus 4.8Anthropic956
8Claude Opus 4.7Anthropic956
9Claude Opus 4.7 (batch)Anthropic956
10Claude Opus 4.8 (batch)Anthropic956
11GPT-5.5 ProOpenAI936
12GPT-5.5 Pro (batch)OpenAI936
13GPT-5.5OpenAI936
14GPT-5.5 (batch)OpenAI936
15Gemini 3.1 Pro Preview Custom ToolsGoogle926
16Gemini 3.1 Pro PreviewGoogle926
17Gemini 3.1 Pro Preview (batch)Google926
18GPT-5.4 ProOpenAI926
19GPT-5.4 Pro (batch)OpenAI926
20GPT-5.4OpenAI926
21GPT-5.4 (batch)OpenAI926
22GPT-5.3-CodexOpenAI916
23GPT-5.2-CodexOpenAI916
24GPT-5.2 ProOpenAI916
25GPT-5.2 Pro (batch)OpenAI916
26GPT-5.2OpenAI916
27GPT-5.2 (batch)OpenAI916
28Claude Opus 4.6Anthropic906
29Claude Opus 4.6 (batch)Anthropic906
30GPT-5.6 Luna ProOpenAI896
31GPT-5.6 Luna Pro (batch)OpenAI896
32GPT-5.6 LunaOpenAI896
33GPT-5.6 Luna (batch)OpenAI896
34GPT-5.6 Terra ProOpenAI896
35GPT-5.6 Terra Pro (batch)OpenAI896
36GPT-5.6 TerraOpenAI896
37GPT-5.6 Terra (batch)OpenAI896
38GPT-5.6 Sol ProOpenAI896
39GPT-5.6 Sol Pro (batch)OpenAI896
40GPT-5.6 SolOpenAI896
41GPT-5.6 Sol (batch)OpenAI896
42Grok 4.7xAI896
43Grok 4.6xAI896
44Grok 4.5xAI896
45Grok 4.3xAI896
46Grok 4.20xAI896
47GPT-5.1-Codex-MaxOpenAI896
48GPT-5.1OpenAI896
49GPT-5.1-CodexOpenAI896
50GPT-5 ProOpenAI896
51GPT-5 Pro (batch)OpenAI896
52GPT-5OpenAI896
53GPT-5 (batch)OpenAI896
54Gemini 3 Flash PreviewGoogle886
55Gemini 3 Flash Preview (batch)Google886
56GPT-5.1 (batch)OpenAI886
57GPT-5.1-Codex-MiniOpenAI886
58o3 ProOpenAI876
59o3OpenAI876
60o3 (batch)OpenAI876
61Claude Sonnet 5Anthropic856
62Claude Sonnet 4.6Anthropic856
63Claude Sonnet 4.6 (batch)Anthropic856
64Claude Opus 4.5Anthropic856
65Claude Opus 4.5 (batch)Anthropic856
66Gemini 2.5 ProGoogle846
67Gemini 2.5 Pro (batch)Google846
68Gemini 2.5 Pro Preview 06-05Google846
69Claude Sonnet 4.5Anthropic826
70Claude Sonnet 4.5 (batch)Anthropic826
71GPT-6 AstraOpenAI826
72GPT-6 Astra (batch)OpenAI826
73GPT-6 Astra ProOpenAI826
74GPT-6 Astra Pro (batch)OpenAI826
75o4 MiniOpenAI816
76o4 Mini (batch)OpenAI816
77Gemini 3.8 FlashGoogle816
78Gemini 3.8 Flash (batch)Google816
79Grok 4.3 (batch)xAI816
80Muse Spark 1.3meta816
81Muse Spark 1.2meta816
82Muse Spark 1.1meta816
83GPT-5.4 NanoOpenAI796
84GPT-5.4 Nano (batch)OpenAI796
85GPT-5.4 MiniOpenAI796
86GPT-5.4 Mini (batch)OpenAI796
87Gemini 3.1 Flash Lite PreviewGoogle796
88Gemini 2.5 Flash LiteGoogle796
89Gemini 2.5 Flash Lite (batch)Google796
90Gemini 2.5 FlashGoogle796
91Gemini 2.5 Flash (batch)Google796
92Gemini 3.5 FlashGoogle796
93Gemini 3.5 Flash (batch)Google796
94GPT-5 Mini (batch)OpenAI776
95GPT-5 Nano (batch)OpenAI776
96Gemini 3.5 Flash LiteGoogle776
97Gemini 3.5 Flash Lite (batch)Google776
98Claude Opus 4.1 (batch)Anthropic756
99o1OpenAI746
100o4 Mini HighOpenAI726
101Claude Haiku 4.5Anthropic706
102Claude Haiku 4.5 (batch)Anthropic706
103GPT-5 NanoOpenAI626
104GPT-5 MiniOpenAI546
105GPT-6 Luna ProOpenAI406
106GPT-6 Luna Pro (batch)OpenAI406
107GPT-6 LunaOpenAI406
108GPT-6 Luna (batch)OpenAI406
109GPT-6 Sol ProOpenAI406
110GPT-6 Sol Pro (batch)OpenAI406
111GPT-6 SolOpenAI406
112GPT-6 Sol (batch)OpenAI406
113Claude Opus 5.5 (batch)Anthropic406
114GPT Astra Latest~openai406
115GPT Sol Latest~openai406
116GPT Terra Latest~openai406
117GPT Luna Latest~openai406
118Fugu Ultra v2sakana406
119Fugu Maxsakana406
120Muse Spark 1.3 Contributormeta406
121Muse Spark 1.2 Contributormeta406
122Gemini 3.7 FlashGoogle406
123Gemini 3.7 Flash (batch)Google406
124Sakana Namazusakana406
125Claude Opus 5 (batch)Anthropic406
126Gemini 3.6 FlashGoogle-6
127Gemini 3.6 Flash (batch)Google-6
128Grok Latest~x-ai-6
129Claude Sonnet 5 (batch)Anthropic-6
130Fugu Ultrasakana-6
131Nano Banana Pro (Gemini 3 Pro Image)Google-6
132Claude Fable Latest~anthropic-6
133Grok Build 0.1xAI-6
134Gemini 3.1 Flash LiteGoogle-6
135Gemini 3.1 Flash Lite (batch)Google-6
136Claude Haiku Latest~anthropic-6
137GPT Mini Latest~openai-6
138Gemini Pro Latest~google-6
139Gemini Flash Latest~google-6
140Claude Sonnet Latest~anthropic-6
141Claude Opus Latest~anthropic-6
142GPT-5.2 ChatOpenAI915
143Gemma 4 31BGoogle815
144Gemma 4 31B (free)Google815
145Qwen3.5 397B A17BAlibaba795
146GLM 5.3 FlashZhipu AI785
147GLM 5.3 FlashXZhipu AI785
148GLM 5.3 Flash (batch)Zhipu AI785
149Qwen3.5-122B-A10BAlibaba785
150Qwen3.5-27BAlibaba775
151MiMo-V2.6-ProXiaomi765
152Qwen3.5-35B-A3BAlibaba765
153Kimi K2.6Moonshot AI765
154Qwen3.8 Max (0902)Alibaba765
155Qwen3.7 PlusAlibaba765
156o3 MiniOpenAI755
157Qwen3.5 Plus 2026-04-20Alibaba755
158Claude Opus 4.1Anthropic745
159o3 Mini (batch)OpenAI745
160Qwen3.6 PlusAlibaba745
161MiniMax M3MiniMax745
162Claude Sonnet 4Anthropic745
163Qwen3.8 27BAlibaba735
164MiMo-V2.5Xiaomi735
165Gemma 4 26B A4B Google735
166Gemma 4 26B A4B (free)Google735
167Qwen3.8 27B (free)Alibaba735
168GLM 5V TurboZhipu AI725
169GPT-4o (batch)OpenAI725
170Mistral Medium 3.5Mistral AI725
171Qwen3 VL 235B A22B ThinkingAlibaba695
172Qwen3.5-FlashAlibaba695
173Qwen3.5 Plus 2026-02-15Alibaba685
174GPT-4.1OpenAI685
175GPT-4.1 (batch)OpenAI685
176Qwen3.5-9BAlibaba675
177GLM 4.6VZhipu AI665
178o3 Mini HighOpenAI645
179GPT-4o-mini (batch)OpenAI635
180GLM 4.5VZhipu AI625
181GPT-4 Turbo (batch)OpenAI625
182GPT-4.1 Mini (batch)OpenAI595
183GPT-4.1 Nano (batch)OpenAI595
184GPT-4.1 MiniOpenAI565
185Command A+Cohere515
186Kimi K2.5Moonshot AI485
187GPT-4.1 NanoOpenAI425
188Ember-1fireworks405
189Qwen3.8 Max PrimeAlibaba405
190MiMo-V2.6-Pro-UltraSpeedXiaomi405
191MiMo-V2.6-FlashXiaomi405
192Qwen3.8 Omni FlashAlibaba405
193Ternary Bonsai 2 27Bprism-ml405
194DeepSeek Flash Latest~deepseek405
195Ling 3.0 Flash VLinclusionai405
196DeepSeek V4.1 FlashDeepSeek405
197DeepSeek V4.1 Flash (batch)DeepSeek405
198GLM Flash Latest~z-ai405
199Qwen3.8 FlashAlibaba405
200DeepSeek V4 Flash Vision ExpDeepSeek405
201Dots3-Note Preview (free)dots-studio405
202Seed 2.1 TurboByteDance405
203Seed-2.0-CodeByteDance405
204Muse Glimmer 30Bmeta405
205Qwen3.7 FlashAlibaba405
206Space Bunny Alphastealth-5
207Kimi K3Moonshot AI-5
208Kimi K3 (batch)Moonshot AI-5
209Kimi K2.7 CodeMoonshot AI-5
210Step 3.7 FlashStepFun-5
211GPT Chat LatestOpenAI-5
212Mistral Medium 3.5 (batch)Mistral AI-5
213Kimi Latest~moonshotai-5
214Qwen3.6 FlashAlibaba-5
215Qwen3.6 35B A3BAlibaba-5
216Qwen3.6 27BAlibaba-5
217Mistral Small 4Mistral AI-5
218Mistral Small 4 (batch)Mistral AI-5
219Seed-2.0-LiteByteDance-5
220Seed-2.0-MiniByteDance-5
221Seed 1.6 FlashByteDance-5
222Seed 1.6ByteDance-5
223Qwen3 VL 8B ThinkingAlibaba-5
224Qwen3 VL 30B A3B ThinkingAlibaba-5

Strong Agent Models

4 agentic capabilities
#ModelScore
1DeepSeek V3.2DeepSeek83
2R1 0528DeepSeek79
3GLM 5.3 PrimeZhipu AI79
4GLM 5.3Zhipu AI79
5GLM 5.3 (batch)Zhipu AI79
6GLM 5.2Zhipu AI79
7GLM 5.1Zhipu AI78
8GLM 5 TurboZhipu AI78
9GLM 5Zhipu AI78
10MiMo-V2.5-ProXiaomi76
11GLM 4.7Zhipu AI75
12GLM 4.6Zhipu AI75
13GLM 4.5Zhipu AI75
14Qwen3.7 MaxAlibaba75
15Qwen3.6 Max PreviewAlibaba75
16Hy3Tencent74
17R1DeepSeek74
18Inklingthinkingmachines73
19Inkling (free)thinkingmachines73
20DeepSeek V3.2 ExpDeepSeek72

What Makes a Model Good for Agents?

Function Calling (Required)

The foundation of agentic AI. Models must be able to decide when and how to call external tools - APIs, databases, file systems, web browsers. Without this, agents can only produce text, not take action.

JSON Mode (Highly Recommended)

Agents need to produce structured, parseable output reliably. JSON mode ensures the model always returns valid JSON, preventing the parsing errors that break automated workflows.

Reasoning (For Complex Tasks)

Chain-of-thought reasoning helps agents plan multi-step sequences, recover from errors, and make better decisions about which tools to use. Critical for complex autonomous workflows.

Streaming (For Real-Time UX)

Streaming lets you observe agent actions in real-time - see which tools it's calling, watch it reason through problems, and intervene early if it goes off track. Essential for interactive agent interfaces.

Vision (For Multimodal Agents)

Vision-capable agents can process screenshots, analyze charts, read documents, and interact with visual interfaces. Required for browser automation, document processing, and GUI agents.

Web Search (For Research Agents)

Built-in web search lets agents find current information without custom search tool integrations. Ideal for research agents, fact-checkers, and competitive analysis workflows.

Frequently Asked Questions

Agent-capable models need strong function calling (tool use), reasoning ability, long context windows, and reliable instruction following. They must decide when to use tools, handle multi-step plans, and recover from errors.

GPT-4o, Claude 3.5 Sonnet, and Gemini 2.0 are the top choices for AI agents. They offer reliable function calling, strong reasoning, and the ability to maintain context across long multi-step workflows.

Chatbots respond to individual messages. AI agents autonomously execute multi-step tasks - browsing the web, writing code, calling APIs, and making decisions. Agents use tool calling and planning capabilities that go beyond simple conversation.

Best AI Models for Building Agents (2026) | LM Market Cap