refactor: extract prelude models to models.yaml (#451)

1 files changed, 324 insertions, 0 deletions

diff --git a/models.yaml b/models.yaml
new file mode 100644
index 0000000..29417ab
--- /dev/null
+++ b/models.yaml
@@ -0,0 +1,324 @@
+- type: openai
+  # docs: 
+  #   - https://platform.openai.com/docs/models
+  #   - https://openai.com/pricing
+  #   - https://platform.openai.com/docs/api-reference/chat
+  models:
+    - name: gpt-3.5-turbo
+      max_input_tokens: 16385
+      max_output_tokens?: 4096
+      input_price: 0.5
+      output_price: 1.5
+    - name: gpt-3.5-turbo-1106
+      max_input_tokens: 16385
+      max_output_tokens?: 4096
+      input_price: 1
+      output_price: 2
+    - name: gpt-4-turbo
+      max_input_tokens: 128000
+      max_output_tokens?: 4096
+      input_price: 10
+      output_price: 30
+      supports_vision: true
+    - name: gpt-4-turbo-preview
+      max_input_tokens: 128000
+      max_output_tokens?: 4096
+      input_price: 10
+      output_price: 30
+    - name: gpt-4-1106-preview
+      max_input_tokens: 128000
+      max_output_tokens?: 4096
+      input_price: 10
+      output_price: 30
+    - name: gpt-4-vision-preview
+      max_input_tokens: 128000
+      max_output_tokens: 4096
+      input_price: 10
+      output_price: 30
+      supports_vision: true
+    - name: gpt-4
+      max_input_tokens: 8192
+      max_output_tokens?: 4096
+      input_price: 30
+      output_price: 60
+    - name: gpt-4-32k
+      max_input_tokens: 32768
+      max_output_tokens?: 4096
+      input_price: 60
+      output_price: 120
+
+- type: gemini 
+  docs:
+    # - https://ai.google.dev/models/gemini
+    # - https://ai.google.dev/pricing
+    # - https://ai.google.dev/api/rest/v1beta/models/streamGenerateContent
+  models:
+    - name: gemini-1.0-pro-latest
+      max_input_tokens: 30720
+      max_output_tokens?: 2048
+      input_price: 0.5
+      output_price: 1.5
+    - name: gemini-1.0-pro-vision-latest
+      max_input_tokens: 12288
+      max_output_tokens?: 4096
+      input_price: 0.5
+      output_price: 1.5
+      supports_vision: true
+    - name: gemini-1.5-pro-latest
+      max_input_tokens: 1048576
+      max_output_tokens?: 8192
+      input_price: 7
+      output_price: 21
+      supports_vision: true
+
+- type: claude
+  # docs:
+  #   - https://docs.anthropic.com/claude/docs/models-overview
+  #   - https://docs.anthropic.com/claude/reference/messages-streaming
+  models:
+    - name: claude-3-opus-20240229
+      max_input_tokens: 200000
+      max_output_tokens: 4096
+      input_price: 15
+      output_price: 75
+      supports_vision: true
+    - name: claude-3-sonnet-20240229
+      max_input_tokens: 200000
+      max_output_tokens: 4096
+      input_price: 3
+      output_price: 15
+      supports_vision: true
+    - name: claude-3-haiku-20240307
+      max_input_tokens: 200000
+      max_output_tokens: 4096
+      input_price: 0.25
+      output_price: 1.25
+      supports_vision: true
+
+- type: mistral
+  # docs:
+  #   - https://docs.mistral.ai/platform/endpoints/
+  #   - https://mistral.ai/technology/#pricing
+  #   - https://docs.mistral.ai/api/
+  models:
+    - name: open-mistral-7b
+      max_input_tokens: 32000
+      max_output_tokens?: 8191
+      input_price: 0.25
+      output_price: 0.25
+    - name: open-mixtral-8x7b
+      max_input_tokens: 32000
+      max_output_tokens?: 8191
+      input_price: 0.7
+      output_price: 0.7
+    - name: open-mixtral-8x22b
+      max_input_tokens: 64000
+      max_output_tokens?: 8191
+      input_price: 2
+      output_price: 6
+    - name: mistral-small-latest
+      max_input_tokens: 32000
+      max_output_tokens?: 8191
+      input_price: 2
+      output_price: 6
+    - name: mistral-medium-latest
+      max_input_tokens: 32000
+      max_output_tokens?: 8191
+      input_price: 2.7
+      output_price: 8.1
+    - name: mistral-large-latest
+      max_input_tokens: 32000
+      max_output_tokens?: 8191
+      input_price: 8
+      output_price: 24
+
+- type: cohere
+  # docs:
+  #   - https://docs.cohere.com/docs/command-r
+  #   - https://docs.cohere.com/docs/command-r-plus
+  #   - https://cohere.com/pricing
+  #   - https://docs.cohere.com/reference/chat
+  models:
+    - name: command-r
+      max_input_tokens: 128000
+      max_output_tokens?: 4096
+      input_price: 0.5
+      output_price: 1.5
+    - name: command-r-plus
+      max_input_tokens: 128000
+      max_output_tokens?: 4096
+      input_price: 3
+      output_price: 15
+
+- type: perplexity
+  # docs:
+  #   - https://docs.perplexity.ai/docs/model-cards
+  #   - https://docs.perplexity.ai/docs/pricing
+  #   - https://docs.perplexity.ai/reference/post_chat_completions
+  models:
+    - name: sonar-small-chat
+      max_input_tokens: 16384
+    - name: sonar-small-online
+      max_input_tokens: 12000
+    - name: sonar-medium-chat
+      max_input_tokens: 16384
+    - name: sonar-medium-online
+      max_input_tokens: 12000
+
+    - name: llama-3-8b-instruct
+      max_input_tokens: 8192
+      max_output_tokens?: 8192
+      input_price: 0.2
+      output_price: 0.2
+    - name: llama-3-70b-instruct
+      max_input_tokens: 8192
+      max_output_tokens?: 8192
+      input_price: 1
+      output_price: 1
+    - name: codellama-70b-instruct
+      max_input_tokens: 16384
+      max_output_tokens?: 4096
+      input_price: 1
+      output_price: 1
+    - name: mistral-7b-instruct
+      max_input_tokens: 16384
+      max_output_tokens?: 8191
+      input_price: 0.2
+      output_price: 0.2
+    - name: mixtral-8x7b-instruct
+      max_input_tokens: 16384
+      max_output_tokens?: 8191
+      input_price: 0.6
+      output_price: 0.6
+    - name: mixtral-8x22b-instruct
+      max_input_tokens: 16384
+      max_output_tokens?: 8191
+      input_price: 1
+      output_price: 1
+
+- type: groq
+  # docs:
+  #   - https://console.groq.com/docs/models
+  #   - https://console.groq.com/docs/text-chat
+  models:
+    - name: llama3-8b-8192
+      max_input_tokens: 8192
+      max_output_tokens: 8192
+    - name: llama3-70b-8192
+      max_input_tokens: 8192
+      max_output_tokens: 8192
+    - name: llama2-70b-4096
+      max_input_tokens: 4096
+      max_output_tokens: 4096
+    - name: mixtral-8x7b-32768
+      max_input_tokens: 32768
+      max_output_tokens: 32768
+    - name: gemma-7b-it
+      max_input_tokens: 8192
+      max_output_tokens: 8192
+
+- type: vertexai
+  # docs:
+  #   - https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models
+  #   - https://cloud.google.com/vertex-ai/generative-ai/pricing
+  #   - https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/gemini
+  models:
+    - name: gemini-1.0-pro
+      max_input_tokens: 24568
+      max_output_tokens: 24568
+      input_price: 0.125
+      output_price: 0.375
+    - name: gemini-1.0-pro-vision
+      max_input_tokens: 14336
+      max_output_tokens: 14336
+      input_price: 0.125
+      output_price: 0.375
+      supports_vision: true
+    - name: gemini-1.5-pro-preview-0409
+      max_input_tokens: 1000000
+      max_output_tokens: 1000000
+      input_price: 2.5
+      output_price: 7.5
+      supports_vision: true
+
+- type: ernie
+  # docs: 
+  #   - https://cloud.baidu.com/doc/WENXINWORKSHOP/s/Nlks5zkzu
+  #   - https://cloud.baidu.com/doc/WENXINWORKSHOP/s/hlrk4akp7
+  models:
+    - name: ernie-4.0-8k-preview
+      max_input_tokens: 5120
+      max_output_tokens: 2048
+      input_price: 16.8
+      output_price: 16.8
+    - name: ernie-3.5-8k-preview
+      max_input_tokens: 5120
+      max_output_tokens: 2048
+      input_price: 1.68
+      output_price: 1.68
+    - name: ernie-speed-128k
+      max_input_tokens: 124000
+      max_output_tokens: 4096
+      input_price: 0.56
+      output_price: 1.12
+    - name: ernie-lite-8k
+      max_input_tokens: 7168
+      max_output_tokens: 2048
+      input_price: 0.42
+      output_price: 0.84
+    - name: ernie-tiny-8k
+      max_input_tokens: 7168
+      max_output_tokens: 2048
+      input_price: 0.14
+      output_price: 0.14
+
+- type: qianwen
+  # docs:
+  #   - https://help.aliyun.com/zh/dashscope/developer-reference/tongyiqianwen-large-language-models/
+  #   - https://help.aliyun.com/zh/dashscope/developer-reference/qwen-vl-plus/
+  models:
+    - name: qwen-turbo
+      max_input_tokens: 6000
+      max_output_tokens?: 1500
+      input_price: 1.12
+      output_price: 1.12
+    - name: qwen-plus
+      max_input_tokens: 30000
+      max_output_tokens?: 2000
+      input_price: 2.8
+      output_price: 2.8
+    - name: qwen-max
+      max_input_tokens: 6000
+      max_output_tokens?: 2000
+      input_price: 16.8
+      output_price: 16.8
+    - name: qwen-max-longcontext
+      max_input_tokens: 28000
+      max_output_tokens?: 2000
+    - name: qwen-vl-plus
+      input_price: 1.12
+      output_price: 1.12
+      supports_vision: true
+    - name: qwen-vl-max
+      input_price: 2.8
+      output_price: 2.8
+      supports_vision: true
+
+- type: moonshot
+  # docs:
+  #   - https://platform.moonshot.cn/docs/intro
+  #   - https://platform.moonshot.cn/docs/pricing
+  #   - https://platform.moonshot.cn/docs/api-reference
+  models:
+    - name: moonshot-v1-8k
+      max_input_tokens: 8000
+      input_price: 1.68
+      output_price: 1.68
+    - name: moonshot-v1-32k
+      max_input_tokens: 32000
+      input_price: 3.36
+      output_price: 3.36
+    - name: moonshot-v1-128k
+      max_input_tokens: 128000
+      input_price: 8.4
+      output_price: 8.4