genai-prices/prices/providers/cerebras.yml at main · pydantic/genai-prices · GitHub

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
167
168
169
170
171
172
173
# yaml-language-server: $schema=.schema.json
name: Cerebras
id: cerebras
pricing_urls:
  - https://www.cerebras.ai/pricing#pricing
  - https://inference-docs.cerebras.ai/models/openai-oss
  - https://inference-docs.cerebras.ai/models/gemma-4-31b
  - https://inference-docs.cerebras.ai/support/change-log
api_pattern: 'https://api\.cerebras\.ai'
model_match:
  contains: cerebras
provider_match:
  contains: cerebras

# Cerebras uses OpenAI-compatible API format
# https://inference-docs.cerebras.ai/api-reference/chat-completions
extractors:
  - api_flavor: chat
    root: usage
    mappings:
      - path: prompt_tokens
        dest: input_tokens
      - path: [completion_tokens_details, reasoning_tokens]
        dest: output_reasoning_tokens
        required: false
      - path: completion_tokens
        dest: output_tokens

models:
  - id: gemma-4-31b
    name: Gemma 4 31B
    description: >-
      Google's Gemma 4 31B open model for chat, instruction following, and multilingual
      tasks, served with fast inference on Cerebras hardware (~1,850 tokens/second).
    match:
      or:
        - equals: gemma-4-31b
        - starts_with: cerebras/gemma-4-31b
        - starts_with: cerebras:gemma-4-31b
    context_window: 131072
    deprecated: true
    prices_checked: 2026-10-15
    price_comments: >-
      Developer tier pricing. Free tier: 65k context, Paid tier: 131k context.
      The Cerebras public models API (https://api.cerebras.ai/public/v1/models, checked 2026-10-15)
      no longer lists this model; prices kept for historical usage.
    prices:
      input_mtok: 0.99
      output_mtok: 1.49

  - id: gpt-oss-120b
    name: GPT-OSS 120B
    description: >-
      OpenAI's flagship open source model, built on a Mixture-of-Experts (MoE) architecture with
      120 billion parameters and 128 experts. Delivers frontier reasoning capabilities with
      record-breaking inference speeds on Cerebras hardware (~3,000 tokens/second).
    match:
      or:
        - equals: gpt-oss-120b
        - starts_with: cerebras/gpt-oss-120b
        - starts_with: cerebras:gpt-oss-120b
    context_window: 131072
    prices_checked: 2026-08-27
    price_comments: "Developer tier pricing. Free tier: 65k context, Paid tier: 131k context."
    prices:
      input_mtok: 0.35
      output_mtok: 0.75

  - id: llama-3.3-70b
    name: Llama 3.3 70B
    description: >-
      Meta's enhanced 70B model delivering 405B-level accuracy. Optimized for chat, coding,
      instruction following, mathematics, and reasoning with high-speed inference on Cerebras
      hardware (~2,100 tokens/second).
    match:
      or:
        - equals: llama-3.3-70b
        - starts_with: cerebras/llama-3.3-70b
        - starts_with: cerebras:llama-3.3-70b
    deprecated: true
    context_window: 128000
    prices_checked: 2025-11-07
    price_comments: No longer listed in the Cerebras Inference docs (checked 2026-07-17 and 2026-08-27); last known rates kept for historical usage.
    prices:
      input_mtok: 0.85
      output_mtok: 1.20

  - id: llama3.1-8b
    name: Llama 3.1 8B
    description: >-
      Meta's Llama 3.1 8B model for general-purpose tasks including chat, coding,
      and instruction following. Optimized for fast inference on Cerebras hardware
      (~2,200 tokens/second).
    match:
      or:
        - equals: llama3.1-8b
        - starts_with: cerebras/llama3.1-8b
        - starts_with: cerebras:llama3.1-8b
    deprecated: true
    context_window: 32768
    prices_checked: 2025-11-07
    price_comments: No longer listed in the Cerebras Inference docs (checked 2026-07-17 and 2026-08-27); last known rates kept for historical usage.
    prices:
      input_mtok: 0.10
      output_mtok: 0.10

  - id: qwen-3-32b
    name: Qwen 3 32B
    description: >-
      Qwen's 32B parameter model with enhanced reasoning and coding capabilities.
      Supports both standard and reasoning modes for complex tasks, with fast inference
      speeds on Cerebras hardware (~2,600 tokens/second).
    match:
      or:
        - equals: qwen-3-32b
        - starts_with: cerebras/qwen-3-32b
        - starts_with: cerebras:qwen-3-32b
    deprecated: true
    context_window: 131072
    prices_checked: 2025-11-07
    price_comments: No longer listed in the Cerebras Inference docs (checked 2026-07-17 and 2026-08-27); last known rates kept for historical usage.
    prices:
      input_mtok: 0.40
      output_mtok: 0.80

  - id: qwen-3-coder-480b
    name: qwen-3-coder-480b
    match:
      equals: qwen-3-coder-480b
    deprecated: true
    prices_checked: 2026-10-09
    # No context_window: Cerebras served this with a tier-dependent window (64k free tier vs 128k paid tiers),
    # so no single value is true; deprecated by Cerebras on 2025-11-05.
    price_comments: >-
      No longer listed in the Cerebras Inference models API
      (https://api.cerebras.ai/public/v1/models, last checked 2026-10-09); kept for historical usage.
    prices: {}

  - id: qwen-3.8-27b
    name: Qwen 3.8 27B
    description: >-
      Qwen's 27B vision-language model for coding, research, and long-horizon agentic tasks.
    match:
      or:
        - equals: qwen-3.8-27b
        - starts_with: cerebras/qwen-3.8-27b
        - starts_with: cerebras:qwen-3.8-27b
    context_window: 65536
    prices_checked: 2026-09-29
    price_comments: Rates and context length from https://api.cerebras.ai/public/v1/models.
    prices:
      input_mtok: 0.99
      output_mtok: 1.49

  - id: zai-glm-4.7
    name: Z.ai GLM 4.7
    description: >-
      Z.ai's GLM 4.7, a 355 billion parameter model with reasoning enabled by default,
      served with fast inference on Cerebras hardware (~1,000 tokens/second).
    match:
      or:
        - equals: zai-glm-4.7
        - starts_with: cerebras/zai-glm-4.7
        - starts_with: cerebras:zai-glm-4.7
    context_window: 131072
    deprecated: true
    prices_checked: 2026-08-27
    price_comments: >-
      Deprecated by Cerebras on 2026-08-17 (see https://inference-docs.cerebras.ai/support/change-log) and no longer
      in the model catalog; last published developer tier rates kept for historical usage.
    prices:
      input_mtok: 2.25
      output_mtok: 2.75