cerebras.json

 1{
 2    "name": "Cerebras",
 3    "id": "cerebras",
 4    "type": "openai",
 5    "api_key": "$CEREBRAS_API_KEY",
 6    "api_endpoint": "https://api.cerebras.ai/v1",
 7    "default_large_model_id": "qwen-3-coder-480b",
 8    "default_small_model_id": "qwen-3-32b",
 9    "models": [
10        {
11            "id": "llama-4-scout-17b-16e-instruct",
12            "name": "Llama 4 Scout",
13            "cost_per_1m_in": 0.65,
14            "cost_per_1m_out": 0.85,
15            "context_window": 32768,
16            "default_max_tokens": 4000,
17            "can_reason": true,
18            "supports_attachments": false
19        },
20        {
21            "id": "llama3.1-8b",
22            "name": "Llama 3.1 8B",
23            "cost_per_1m_in": 0.1,
24            "cost_per_1m_out": 0.1,
25            "context_window": 32768,
26            "default_max_tokens": 4000,
27            "can_reason": true,
28            "supports_attachments": false
29        },
30        {
31            "id": "llama-3.3-70b",
32            "name": "Llama 3.3 70B",
33            "cost_per_1m_in": 0.85,
34            "cost_per_1m_out": 1.2,
35            "context_window": 128000,
36            "default_max_tokens": 4000,
37            "can_reason": true,
38            "supports_attachments": false
39        },
40        {
41            "id": "qwen-3-32b",
42            "name": "Qwen 3 32B",
43            "cost_per_1m_in": 0.4,
44            "cost_per_1m_out": 0.8,
45            "context_window": 128000,
46            "default_max_tokens": 32768,
47            "can_reason": true,
48            "supports_attachments": false
49        },
50        {
51            "id": "llama-4-maverick-17b-128e-instruct",
52            "name": "Llama 4 Maverick",
53            "cost_per_1m_in": 0.2,
54            "cost_per_1m_out": 0.6,
55            "context_window": 32768,
56            "default_max_tokens": 4000,
57            "can_reason": true,
58            "supports_attachments": false
59        },
60        {
61            "id": "qwen-3-235b-a22b-instruct-2507",
62            "name": "Qwen 3 235B Instruct",
63            "cost_per_1m_in": 0.6,
64            "cost_per_1m_out": 1.2,
65            "context_window": 131072,
66            "default_max_tokens": 16384,
67            "can_reason": true,
68            "supports_attachments": false
69        },
70        {
71            "id": "qwen-3-235b-a22b-thinking-2507",
72            "name": "Qwen 3 235B Thinking",
73            "cost_per_1m_in": 0.6,
74            "cost_per_1m_out": 1.2,
75            "context_window": 128000,
76            "default_max_tokens": 32768,
77            "can_reason": true,
78            "supports_attachments": false
79        },
80        {
81            "id": "qwen-3-coder-480b",
82            "name": "Qwen 3 480B Coder",
83            "cost_per_1m_in": 2.0,
84            "cost_per_1m_out": 2.0,
85            "context_window": 131072,
86            "default_max_tokens": 65536,
87            "can_reason": true,
88            "supports_attachments": false
89        }
90    ]
91}