Repository navigation
Expand file tree
/
Copy pathmodel_list.txt
More file actions
166 lines (138 loc) · 6.86 KB
/
Copy pathmodel_list.txt
File metadata and controls
166 lines (138 loc) · 6.86 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
AutoCua - Supported Providers & Models
========================================
In main.py, pass them straight to run_agent:
run_agent(
mode="computer use",
provider="<provider-name>",
model="<model-name>",
task="...",
)
Copy the names exactly as written below (case-sensitive).
`thinking` is the reasoning level the agents send; `coder` is the level
Coder mode sends instead (set per model as coder_effort in each
provider's view.py, next to the normal one).
A name that is not on this list is NOT validated - it is sent to the
provider verbatim and comes back as a 404 model_not_found.
--------------------------------------------------------------
provider="openrouter"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
gemini-3.6-flash minimal low yes
gemini-3.8-flash low medium yes
gemini-4-argon low medium yes
gpt-6-astra medium high yes
gpt-6-luna low medium yes
gpt-6-sol low medium yes
gpt-6.1-terra low medium yes
claude-opus-5 medium high yes
claude-sonnet-5 low medium yes
claude-sonnet-5.5 low medium yes
claude-opus-5.5 medium high yes
grok-4.7 low medium yes
mimo-v2.6-pro-ultraspeed low low yes
deepseek-v4.1-flash low medium yes
deepseek-v4.1-flash-1017 low low yes
minimax-m3.1-flash low medium yes
muse-wonder low medium yes
claude-opus-5-fast low medium yes
mistral-medium-3.5 none none yes
qwen3.8-max-0902 low medium yes
qwen3.8-flash low medium yes
--------------------------------------------------------------
provider="groq"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
qwen3.8-27b - medium yes
--------------------------------------------------------------
provider="openai"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
gpt-6-astra medium high yes
--------------------------------------------------------------
provider="anthropic"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
claude-haiku-4.5 sampling sampling yes
claude-opus-5 low medium yes
claude-sonnet-5 low medium yes
claude-sonnet-5.5 low medium yes
claude-opus-5.5 low medium yes
--------------------------------------------------------------
provider="google"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
gemini-4-argon low medium yes
gemini-3.6-flash low medium yes
gemini-3.8-flash low medium yes
gemini-4-argon-vertex low medium yes
gemini-3.6-flash-vertex low medium yes
gemini-3.8-flash-vertex low medium yes
--------------------------------------------------------------
provider="perplexity"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
gemini-4-argon low medium yes
gemini-3.6-flash low medium yes
gemini-3.8-flash low medium yes
grok-4.5 low medium yes
kimi-k3 low medium yes
glm-5.3-flash low high yes
--------------------------------------------------------------
provider="together"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
minimax-m3 default default yes
inkling default default yes
muse-glimmer-30b default default yes
--------------------------------------------------------------
provider="cerebras"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
qwen-3.8-27b low high yes
--------------------------------------------------------------
provider="requesty"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
claude-sonnet-5 low medium yes
claude-sonnet-5.5 low medium yes
claude-opus-5 medium high yes
claude-opus-5.5 medium high yes
gemini-3.6-flash default default yes
gemini-4-argon low medium yes
gpt-5.4 default default yes
grok-4.7 low medium yes
particle-1 low medium yes
minimax-m3.1-flash low medium yes
deepseek-v4.1-flash-1017 none none yes
--------------------------------------------------------------
provider="aws"
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
kimi-k3 low medium yes
grok-4.7 low medium yes
claude-opus-4.7 low medium yes
claude-opus-4.6 low medium yes
claude-sonnet-4.6 low medium yes
--------------------------------------------------------------
provider="ollama" (on this machine: no API key; OLLAMA_HOST to point
elsewhere. Any model `ollama list` shows works if it
has tools and vision in `ollama show <model>`)
--------------------------------------------------------------
model name thinking coder vision
--------------------------------------------------------------
qwen3.8:27b none medium yes
qwen3.6:35b-a3b-nvfp4 none medium yes
ornith-1.5:35b none medium yes
--------------------------------------------------------------
63 models across 11 providers.
--------------------------------------------------------------