azure.json

  1{
  2  "name": "Azure OpenAI",
  3  "id": "azure",
  4  "type": "azure",
  5  "api_key": "$AZURE_OPENAI_API_KEY",
  6  "api_endpoint": "$AZURE_OPENAI_API_ENDPOINT",
  7  "default_large_model_id": "gpt-5",
  8  "default_small_model_id": "gpt-5-mini",
  9  "models": [
 10    {
 11      "id": "gpt-5",
 12      "name": "GPT-5",
 13      "cost_per_1m_in": 1.25,
 14      "cost_per_1m_out": 10,
 15      "cost_per_1m_in_cached": 0.25,
 16      "cost_per_1m_out_cached": 0.25,
 17      "context_window": 400000,
 18      "default_max_tokens": 128000,
 19      "can_reason": true,
 20      "reasoning_levels": [
 21        "minimal",
 22        "low",
 23        "medium",
 24        "high"
 25      ],
 26      "default_reasoning_effort": "minimal",
 27      "supports_attachments": true
 28    },
 29    {
 30      "id": "gpt-5-mini",
 31      "name": "GPT-5 Mini",
 32      "cost_per_1m_in": 0.25,
 33      "cost_per_1m_out": 2,
 34      "cost_per_1m_in_cached": 0.025,
 35      "cost_per_1m_out_cached": 0.025,
 36      "context_window": 400000,
 37      "default_max_tokens": 128000,
 38      "can_reason": true,
 39      "reasoning_levels": [
 40        "low",
 41        "medium",
 42        "high"
 43      ],
 44      "default_reasoning_effort": "low",
 45      "supports_attachments": true
 46    },
 47    {
 48      "id": "gpt-5-nano",
 49      "name": "GPT-5 Nano",
 50      "cost_per_1m_in": 0.05,
 51      "cost_per_1m_out": 0.4,
 52      "cost_per_1m_in_cached": 0.005,
 53      "cost_per_1m_out_cached": 0.005,
 54      "context_window": 400000,
 55      "default_max_tokens": 128000,
 56      "can_reason": true,
 57      "reasoning_levels": [
 58        "low",
 59        "medium",
 60        "high"
 61      ],
 62      "default_reasoning_effort": "low",
 63      "supports_attachments": true
 64    },
 65    {
 66      "id": "codex-mini-latest",
 67      "name": "Codex Mini",
 68      "cost_per_1m_in": 1.5,
 69      "cost_per_1m_out": 6,
 70      "cost_per_1m_in_cached": 0,
 71      "cost_per_1m_out_cached": 0.375,
 72      "context_window": 200000,
 73      "default_max_tokens": 50000,
 74      "can_reason": true,
 75      "reasoning_levels": [
 76        "low",
 77        "medium",
 78        "high"
 79      ],
 80      "default_reasoning_effort": "medium",
 81      "supports_attachments": true
 82    },
 83    {
 84      "id": "o4-mini",
 85      "name": "o4 Mini",
 86      "cost_per_1m_in": 1.1,
 87      "cost_per_1m_out": 4.4,
 88      "cost_per_1m_in_cached": 0,
 89      "cost_per_1m_out_cached": 0.275,
 90      "context_window": 200000,
 91      "default_max_tokens": 50000,
 92      "can_reason": true,
 93      "has_reasoning_efforts": true,
 94      "default_reasoning_effort": "medium",
 95      "supports_attachments": true
 96    },
 97    {
 98      "id": "o3",
 99      "name": "o3",
100      "cost_per_1m_in": 2,
101      "cost_per_1m_out": 8,
102      "cost_per_1m_in_cached": 0,
103      "cost_per_1m_out_cached": 0.5,
104      "context_window": 200000,
105      "default_max_tokens": 50000,
106      "can_reason": true,
107      "reasoning_levels": [
108        "low",
109        "medium",
110        "high"
111      ],
112      "default_reasoning_effort": "medium",
113      "supports_attachments": true
114    },
115    {
116      "id": "o3-pro",
117      "name": "o3 Pro",
118      "cost_per_1m_in": 20,
119      "cost_per_1m_out": 80,
120      "cost_per_1m_in_cached": 0,
121      "cost_per_1m_out_cached": 0,
122      "context_window": 200000,
123      "default_max_tokens": 50000,
124      "can_reason": true,
125      "reasoning_levels": [
126        "low",
127        "medium",
128        "high"
129      ],
130      "default_reasoning_effort": "medium",
131      "supports_attachments": true
132    },
133    {
134      "id": "gpt-4.1",
135      "name": "GPT-4.1",
136      "cost_per_1m_in": 2,
137      "cost_per_1m_out": 8,
138      "cost_per_1m_in_cached": 0,
139      "cost_per_1m_out_cached": 0.5,
140      "context_window": 1047576,
141      "default_max_tokens": 50000,
142      "can_reason": false,
143      "supports_attachments": true
144    },
145    {
146      "id": "gpt-4.1-mini",
147      "name": "GPT-4.1 Mini",
148      "cost_per_1m_in": 0.39999999999999997,
149      "cost_per_1m_out": 1.5999999999999999,
150      "cost_per_1m_in_cached": 0,
151      "cost_per_1m_out_cached": 0.09999999999999999,
152      "context_window": 1047576,
153      "default_max_tokens": 50000,
154      "can_reason": false,
155      "supports_attachments": true
156    },
157    {
158      "id": "gpt-4.1-nano",
159      "name": "GPT-4.1 Nano",
160      "cost_per_1m_in": 0.09999999999999999,
161      "cost_per_1m_out": 0.39999999999999997,
162      "cost_per_1m_in_cached": 0,
163      "cost_per_1m_out_cached": 0.024999999999999998,
164      "context_window": 1047576,
165      "default_max_tokens": 50000,
166      "can_reason": false,
167      "supports_attachments": true
168    },
169    {
170      "id": "gpt-4.5-preview",
171      "name": "GPT-4.5 (Preview)",
172      "cost_per_1m_in": 75,
173      "cost_per_1m_out": 150,
174      "cost_per_1m_in_cached": 0,
175      "cost_per_1m_out_cached": 37.5,
176      "context_window": 128000,
177      "default_max_tokens": 50000,
178      "can_reason": false,
179      "supports_attachments": true
180    },
181    {
182      "id": "o3-mini",
183      "name": "o3 Mini",
184      "cost_per_1m_in": 1.1,
185      "cost_per_1m_out": 4.4,
186      "cost_per_1m_in_cached": 0,
187      "cost_per_1m_out_cached": 0.55,
188      "context_window": 200000,
189      "default_max_tokens": 50000,
190      "can_reason": true,
191      "reasoning_levels": [
192        "low",
193        "medium",
194        "high"
195      ],
196      "default_reasoning_effort": "medium",
197      "supports_attachments": false
198    },
199    {
200      "id": "gpt-4o",
201      "name": "GPT-4o",
202      "cost_per_1m_in": 2.5,
203      "cost_per_1m_out": 10,
204      "cost_per_1m_in_cached": 0,
205      "cost_per_1m_out_cached": 1.25,
206      "context_window": 128000,
207      "default_max_tokens": 20000,
208      "can_reason": false,
209      "supports_attachments": true
210    },
211    {
212      "id": "gpt-4o-mini",
213      "name": "GPT-4o-mini",
214      "cost_per_1m_in": 0.15,
215      "cost_per_1m_out": 0.6,
216      "cost_per_1m_in_cached": 0,
217      "cost_per_1m_out_cached": 0.075,
218      "context_window": 128000,
219      "default_max_tokens": 20000,
220      "can_reason": false,
221      "reasoning_effort": "",
222      "supports_attachments": true
223    }
224  ]
225}