diff --git a/models/alibaba/qwen3.8-omni-flash.yaml b/models/alibaba/qwen3.8-omni-flash.yaml new file mode 100644 index 00000000..e7bff340 --- /dev/null +++ b/models/alibaba/qwen3.8-omni-flash.yaml @@ -0,0 +1,55 @@ +# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json +provider: alibaba +authType: api_key +apiSurface: openai-chat-completions +model: qwen3.8-omni-flash +params: + - path: max_completion_tokens + type: integer + label: Max tokens + description: Maximum number of tokens to generate, including both reasoning and the final answer. + range: + min: 1 + group: generation_length + - path: temperature + type: number + label: Temperature + description: Controls randomness. Lower values make outputs more focused; higher values make them more varied. + range: + min: 0 + max: 2 + step: 0.1 + group: sampling + - path: top_p + type: number + label: Top P + description: Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. + range: + min: 0 + max: 1 + step: 0.01 + group: sampling + - path: extra_body.top_k + type: integer + label: Top K + description: Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling. + default: 20 + range: + min: 0 + group: sampling + - path: extra_body.enable_thinking + type: boolean + label: Enable thinking + description: Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients. + default: true + group: reasoning + - path: extra_body.thinking_budget + type: integer + label: Thinking budget + description: Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length. + range: + min: 1 + group: reasoning + applicability: + only: + extra_body.enable_thinking: true diff --git a/packages/modelparams-python/src/modelparams/_generated/catalog.json b/packages/modelparams-python/src/modelparams/_generated/catalog.json index 187e2de4..47331631 100644 --- a/packages/modelparams-python/src/modelparams/_generated/catalog.json +++ b/packages/modelparams-python/src/modelparams/_generated/catalog.json @@ -3014,6 +3014,82 @@ } ] }, + { + "provider": "alibaba", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "qwen3.8-omni-flash", + "params": [ + { + "path": "max_completion_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate, including both reasoning and the final answer.", + "group": "generation_length", + "type": "integer", + "range": { + "min": 1 + } + }, + { + "path": "temperature", + "label": "Temperature", + "description": "Controls randomness. Lower values make outputs more focused; higher values make them more varied.", + "group": "sampling", + "type": "number", + "range": { + "min": 0, + "max": 2, + "step": 0.1 + } + }, + { + "path": "top_p", + "label": "Top P", + "description": "Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.", + "group": "sampling", + "type": "number", + "range": { + "min": 0, + "max": 1, + "step": 0.01 + } + }, + { + "path": "extra_body.top_k", + "label": "Top K", + "description": "Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.", + "group": "sampling", + "type": "integer", + "default": 20, + "range": { + "min": 0 + } + }, + { + "path": "extra_body.enable_thinking", + "label": "Enable thinking", + "description": "Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.", + "group": "reasoning", + "type": "boolean", + "default": true + }, + { + "path": "extra_body.thinking_budget", + "label": "Thinking budget", + "description": "Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.", + "group": "reasoning", + "applicability": { + "only": { + "extra_body.enable_thinking": true + } + }, + "type": "integer", + "range": { + "min": 1 + } + } + ] + }, { "provider": "alibaba", "authType": "api_key", diff --git a/packages/modelparams-python/src/modelparams/_generated/model_ids.py b/packages/modelparams-python/src/modelparams/_generated/model_ids.py index 2feab6ff..f6ddf0c1 100644 --- a/packages/modelparams-python/src/modelparams/_generated/model_ids.py +++ b/packages/modelparams-python/src/modelparams/_generated/model_ids.py @@ -48,6 +48,7 @@ "alibaba/qwen3.8-flash", "alibaba/qwen3.8-max", "alibaba/qwen3.8-max-0902", + "alibaba/qwen3.8-omni-flash", "alibaba/qwq-plus", "anthropic/claude-3-5-haiku-20241022", "anthropic/claude-3-5-haiku-latest", @@ -465,6 +466,7 @@ "alibaba/qwen3.8-flash", "alibaba/qwen3.8-max", "alibaba/qwen3.8-max-0902", + "alibaba/qwen3.8-omni-flash", "alibaba/qwq-plus", "anthropic/claude-3-5-haiku-20241022", "anthropic/claude-3-5-haiku-latest", diff --git a/packages/modelparams-python/src/modelparams/_generated/registry.py b/packages/modelparams-python/src/modelparams/_generated/registry.py index a28a37e7..46b802b3 100644 --- a/packages/modelparams-python/src/modelparams/_generated/registry.py +++ b/packages/modelparams-python/src/modelparams/_generated/registry.py @@ -73,6 +73,7 @@ "alibaba/qwen3.8-flash": alibaba.Qwen3_8_FlashParams, "alibaba/qwen3.8-max": alibaba.Qwen3_8_MaxParams, "alibaba/qwen3.8-max-0902": alibaba.Qwen3_8_Max_0902Params, + "alibaba/qwen3.8-omni-flash": alibaba.Qwen3_8_Omni_FlashParams, "alibaba/qwq-plus": alibaba.Qwq_PlusParams, "anthropic/claude-3-5-haiku-20241022": anthropic.Claude_3_5_Haiku_20241022Params, "anthropic/claude-3-5-haiku-latest": anthropic.Claude_3_5_Haiku_LatestParams, diff --git a/packages/modelparams-python/src/modelparams/types/alibaba.py b/packages/modelparams-python/src/modelparams/types/alibaba.py index 2359a1c8..d2e79b3d 100644 --- a/packages/modelparams-python/src/modelparams/types/alibaba.py +++ b/packages/modelparams-python/src/modelparams/types/alibaba.py @@ -602,6 +602,20 @@ ) setattr(Qwen3_8_Max_0902Params, "__pydantic_config__", _PARAMS_CONFIG) +Qwen3_8_Omni_FlashParams = TypedDict( + "Qwen3_8_Omni_FlashParams", + { + "max_completion_tokens": Annotated[int, Field(ge=1)], + "temperature": Annotated[float, Field(ge=0, le=2)], + "top_p": Annotated[float, Field(ge=0, le=1)], + "extra_body.top_k": Annotated[int, Field(ge=0)], + "extra_body.enable_thinking": bool, + "extra_body.thinking_budget": Annotated[int, Field(ge=1)], + }, + total=False, +) +setattr(Qwen3_8_Omni_FlashParams, "__pydantic_config__", _PARAMS_CONFIG) + Qwq_PlusParams = TypedDict( "Qwq_PlusParams", { @@ -659,5 +673,6 @@ "Qwen3_8_FlashParams", "Qwen3_8_MaxParams", "Qwen3_8_Max_0902Params", + "Qwen3_8_Omni_FlashParams", "Qwq_PlusParams", ] diff --git a/packages/modelparams/src/generated/data.ts b/packages/modelparams/src/generated/data.ts index 7e2b21eb..22bc7986 100644 --- a/packages/modelparams/src/generated/data.ts +++ b/packages/modelparams/src/generated/data.ts @@ -3019,6 +3019,82 @@ const GENERATED_CATALOG = [ } ] }, + { + "provider": "alibaba", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "qwen3.8-omni-flash", + "params": [ + { + "path": "max_completion_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate, including both reasoning and the final answer.", + "group": "generation_length", + "type": "integer", + "range": { + "min": 1 + } + }, + { + "path": "temperature", + "label": "Temperature", + "description": "Controls randomness. Lower values make outputs more focused; higher values make them more varied.", + "group": "sampling", + "type": "number", + "range": { + "min": 0, + "max": 2, + "step": 0.1 + } + }, + { + "path": "top_p", + "label": "Top P", + "description": "Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.", + "group": "sampling", + "type": "number", + "range": { + "min": 0, + "max": 1, + "step": 0.01 + } + }, + { + "path": "extra_body.top_k", + "label": "Top K", + "description": "Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.", + "group": "sampling", + "type": "integer", + "default": 20, + "range": { + "min": 0 + } + }, + { + "path": "extra_body.enable_thinking", + "label": "Enable thinking", + "description": "Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.", + "group": "reasoning", + "type": "boolean", + "default": true + }, + { + "path": "extra_body.thinking_budget", + "label": "Thinking budget", + "description": "Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.", + "group": "reasoning", + "applicability": { + "only": { + "extra_body.enable_thinking": true + } + }, + "type": "integer", + "range": { + "min": 1 + } + } + ] + }, { "provider": "alibaba", "authType": "api_key", diff --git a/packages/modelparams/src/generated/defaults.ts b/packages/modelparams/src/generated/defaults.ts index 06302061..30971963 100644 --- a/packages/modelparams/src/generated/defaults.ts +++ b/packages/modelparams/src/generated/defaults.ts @@ -178,6 +178,10 @@ export const DEFAULTS = { "extra_body.top_k": 20, "extra_body.enable_thinking": true, }, + "alibaba/qwen3.8-omni-flash": { + "extra_body.top_k": 20, + "extra_body.enable_thinking": true, + }, "alibaba/qwq-plus": { "extra_body.top_k": 20, }, diff --git a/packages/modelparams/src/generated/model-ids.ts b/packages/modelparams/src/generated/model-ids.ts index 991870c4..c3236d34 100644 --- a/packages/modelparams/src/generated/model-ids.ts +++ b/packages/modelparams/src/generated/model-ids.ts @@ -46,6 +46,7 @@ export const MODEL_IDS = [ "alibaba/qwen3.8-flash", "alibaba/qwen3.8-max", "alibaba/qwen3.8-max-0902", + "alibaba/qwen3.8-omni-flash", "alibaba/qwq-plus", "anthropic/claude-3-5-haiku-20241022", "anthropic/claude-3-5-haiku-latest", diff --git a/packages/modelparams/src/generated/params-by-id.ts b/packages/modelparams/src/generated/params-by-id.ts index 030c78c9..43cd3505 100644 --- a/packages/modelparams/src/generated/params-by-id.ts +++ b/packages/modelparams/src/generated/params-by-id.ts @@ -335,6 +335,14 @@ export type ParamsById = { "extra_body.enable_thinking": boolean; "extra_body.thinking_budget": number; }; + "alibaba/qwen3.8-omni-flash": { + max_completion_tokens: number; + temperature: number; + top_p: number; + "extra_body.top_k": number; + "extra_body.enable_thinking": boolean; + "extra_body.thinking_budget": number; + }; "alibaba/qwq-plus": { max_tokens: number; temperature: number;