diff --git a/models/fireworks/minimax-m3p1-flash.yaml b/models/fireworks/minimax-m3p1-flash.yaml new file mode 100644 index 00000000..3394c415 --- /dev/null +++ b/models/fireworks/minimax-m3p1-flash.yaml @@ -0,0 +1,66 @@ +# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json +provider: fireworks +authType: api_key +apiSurface: openai-chat-completions +model: minimax-m3p1-flash +wireId: accounts/fireworks/models/minimax-m3p1-flash +params: + - path: max_tokens + type: integer + label: Max tokens + description: Maximum number of tokens to generate in the response. + 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: top_k + type: integer + label: Top K + description: Limits token sampling to the top K most likely next tokens. + range: + min: 1 + max: 100 + group: sampling + - path: presence_penalty + type: number + label: Presence penalty + description: Penalizes tokens that have already appeared to encourage a wider variety of content. + range: + min: -2 + max: 2 + step: 0.1 + group: sampling + - path: frequency_penalty + type: number + label: Frequency penalty + description: Penalizes tokens in proportion to how often they have already appeared. + range: + min: -2 + max: 2 + step: 0.1 + group: sampling + - path: response_format.type + type: enum + label: Response format + description: Choose plain text or force valid JSON output. + values: + - text + - json_object + group: output_format diff --git a/models/minimax/MiniMax-M2.1-highspeed.yaml b/models/minimax/MiniMax-M2.1-highspeed.yaml index 5d3934b2..06bc4b2b 100644 --- a/models/minimax/MiniMax-M2.1-highspeed.yaml +++ b/models/minimax/MiniMax-M2.1-highspeed.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2.1-highspeed +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/minimax/MiniMax-M2.1.yaml b/models/minimax/MiniMax-M2.1.yaml index c54a1330..9d1c48da 100644 --- a/models/minimax/MiniMax-M2.1.yaml +++ b/models/minimax/MiniMax-M2.1.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2.1 +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/minimax/MiniMax-M2.5-highspeed.yaml b/models/minimax/MiniMax-M2.5-highspeed.yaml index 09c52065..6412de35 100644 --- a/models/minimax/MiniMax-M2.5-highspeed.yaml +++ b/models/minimax/MiniMax-M2.5-highspeed.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2.5-highspeed +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/minimax/MiniMax-M2.5.yaml b/models/minimax/MiniMax-M2.5.yaml index fa257816..c298bf09 100644 --- a/models/minimax/MiniMax-M2.5.yaml +++ b/models/minimax/MiniMax-M2.5.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2.5 +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/minimax/MiniMax-M2.7-highspeed.yaml b/models/minimax/MiniMax-M2.7-highspeed.yaml index 0e94ab25..28e4afff 100644 --- a/models/minimax/MiniMax-M2.7-highspeed.yaml +++ b/models/minimax/MiniMax-M2.7-highspeed.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2.7-highspeed +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/minimax/MiniMax-M2.7.yaml b/models/minimax/MiniMax-M2.7.yaml index 6e94311f..b5968835 100644 --- a/models/minimax/MiniMax-M2.7.yaml +++ b/models/minimax/MiniMax-M2.7.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2.7 +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/minimax/MiniMax-M2.yaml b/models/minimax/MiniMax-M2.yaml index 939f175d..95050189 100644 --- a/models/minimax/MiniMax-M2.yaml +++ b/models/minimax/MiniMax-M2.yaml @@ -3,6 +3,7 @@ provider: minimax authType: api_key apiSurface: openai-chat-completions model: MiniMax-M2 +status: retired params: - path: max_completion_tokens type: integer diff --git a/models/mistral/zai-glm-5-3-flash.yaml b/models/mistral/zai-glm-5-3-flash.yaml new file mode 100644 index 00000000..2628294a --- /dev/null +++ b/models/mistral/zai-glm-5-3-flash.yaml @@ -0,0 +1,72 @@ +# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json +provider: mistral +authType: api_key +apiSurface: openai-chat-completions +model: zai-glm-5-3-flash +params: + - path: max_tokens + type: integer + label: Max tokens + description: Maximum number of tokens to generate in the completion. + 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: 1.5 + 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. + default: 1 + range: + min: 0.01 + max: 1 + step: 0.01 + group: sampling + - path: random_seed + type: integer + label: Random seed + description: Seed used for deterministic sampling when reproducible outputs are desired. + range: + min: 0 + group: sampling + - path: presence_penalty + type: number + label: Presence penalty + description: Penalizes repeated words or phrases to encourage a wider variety of generated content. + default: 0 + range: + min: -2 + max: 2 + step: 0.1 + group: sampling + - path: frequency_penalty + type: number + label: Frequency penalty + description: Penalizes words based on how often they already appear in the generated text. + default: 0 + range: + min: -2 + max: 2 + step: 0.1 + group: sampling + - path: prompt_mode + type: enum + label: Prompt mode + description: Enables Mistral's reasoning system prompt; leave unset to disable the default reasoning behavior. + values: + - reasoning + group: reasoning + - path: safe_prompt + type: boolean + label: Safe prompt + description: Controls whether Mistral injects its safety prompt before the conversation. + default: false + group: provider_metadata diff --git a/models/moonshot/kimi-k3.1-pro-highspeed.yaml b/models/moonshot/kimi-k3.1-pro-highspeed.yaml new file mode 100644 index 00000000..555ac691 --- /dev/null +++ b/models/moonshot/kimi-k3.1-pro-highspeed.yaml @@ -0,0 +1,73 @@ +# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json +provider: moonshot +authType: api_key +apiSurface: openai-chat-completions +model: kimi-k3.1-pro-highspeed +status: active +params: + - path: max_completion_tokens + type: integer + label: Max tokens + description: Maximum number of tokens to generate in the chat completion. + range: + min: 1 + group: generation_length + - path: thinking.type + type: enum + label: Thinking mode + description: >- + Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; + set disabled to respond directly. + default: enabled + values: + - enabled + - disabled + group: reasoning + - path: thinking.keep + type: enum + label: Keep prior reasoning + description: >- + Set to all to carry reasoning content from earlier assistant turns into the request; leave + null to send only the final answers. + default: null + values: + - all + - null + group: reasoning + - path: response_format.type + type: enum + label: Response format + description: Forces the response into plain text or a JSON object. + default: text + values: + - text + - json_object + group: output_format + - path: tool_choice + type: enum + label: Tool choice + description: >- + Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects + required while thinking is enabled, which is the default on this model. + default: auto + values: + - auto + - none + group: tooling + - path: logprobs + type: boolean + label: Log probabilities + description: Controls whether the response includes log probabilities for the generated tokens. + default: false + group: observability + - path: top_logprobs + type: integer + label: Top log probabilities + description: Number of most likely tokens to return a log probability for at each position. + range: + min: 0 + max: 20 + group: observability + applicability: + only: + logprobs: true diff --git a/models/moonshot/kimi-k3.1-pro.yaml b/models/moonshot/kimi-k3.1-pro.yaml new file mode 100644 index 00000000..4779a926 --- /dev/null +++ b/models/moonshot/kimi-k3.1-pro.yaml @@ -0,0 +1,73 @@ +# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json +provider: moonshot +authType: api_key +apiSurface: openai-chat-completions +model: kimi-k3.1-pro +status: active +params: + - path: max_completion_tokens + type: integer + label: Max tokens + description: Maximum number of tokens to generate in the chat completion. + range: + min: 1 + group: generation_length + - path: thinking.type + type: enum + label: Thinking mode + description: >- + Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; + set disabled to respond directly. + default: enabled + values: + - enabled + - disabled + group: reasoning + - path: thinking.keep + type: enum + label: Keep prior reasoning + description: >- + Set to all to carry reasoning content from earlier assistant turns into the request; leave + null to send only the final answers. + default: null + values: + - all + - null + group: reasoning + - path: response_format.type + type: enum + label: Response format + description: Forces the response into plain text or a JSON object. + default: text + values: + - text + - json_object + group: output_format + - path: tool_choice + type: enum + label: Tool choice + description: >- + Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects + required while thinking is enabled, which is the default on this model. + default: auto + values: + - auto + - none + group: tooling + - path: logprobs + type: boolean + label: Log probabilities + description: Controls whether the response includes log probabilities for the generated tokens. + default: false + group: observability + - path: top_logprobs + type: integer + label: Top log probabilities + description: Number of most likely tokens to return a log probability for at each position. + range: + min: 0 + max: 20 + group: observability + applicability: + only: + logprobs: true diff --git a/models/nvidia/kimi-k3.yaml b/models/nvidia/kimi-k3.yaml index 824087bd..c33eae06 100644 --- a/models/nvidia/kimi-k3.yaml +++ b/models/nvidia/kimi-k3.yaml @@ -4,6 +4,7 @@ authType: api_key apiSurface: openai-chat-completions model: kimi-k3 wireId: moonshotai/kimi-k3 +status: retired params: - path: temperature type: number diff --git a/models/nvidia/nemotron-3.5-bolt-e4b.yaml b/models/nvidia/nemotron-3.5-bolt-e4b.yaml new file mode 100644 index 00000000..d1944896 --- /dev/null +++ b/models/nvidia/nemotron-3.5-bolt-e4b.yaml @@ -0,0 +1,51 @@ +# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json +provider: nvidia +authType: api_key +apiSurface: openai-chat-completions +model: nemotron-3.5-bolt-e4b +wireId: nvidia/nemotron-3.5-bolt-e4b +params: + - path: temperature + type: number + label: Temperature + description: Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call. + default: 1 + range: + max: 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. Not recommended to modify both temperature and top_p in the same call. + default: 0.95 + range: + max: 1 + group: sampling + - path: max_tokens + type: integer + label: Max tokens + description: Maximum number of tokens to generate. Generation stops when this limit is reached. + default: 16384 + range: + min: 1 + max: 32768 + group: generation_length + - path: reasoning_effort + type: enum + label: Reasoning effort + description: Controls the reasoning mode. 'none' disables reasoning tokens, 'low' enables low-effort reasoning, and 'high' enables full reasoning. + default: high + values: + - none + - low + - high + group: reasoning + - path: reasoning_budget + type: integer + label: Reasoning budget + description: Maximum number of tokens the model may use for internal reasoning before being forced to end the reasoning trace. Use -1 to disable budget enforcement. + default: 16384 + range: + min: -1 + max: 32768 + group: reasoning diff --git a/models/openai/gpt-5.5-subscription.yaml b/models/openai/gpt-5.5-subscription.yaml index 58ba16eb..9884b885 100644 --- a/models/openai/gpt-5.5-subscription.yaml +++ b/models/openai/gpt-5.5-subscription.yaml @@ -3,7 +3,7 @@ provider: openai authType: subscription apiSurface: openai-responses model: gpt-5.5 -status: active +status: retired params: - path: reasoning.effort type: enum diff --git a/packages/modelparams-python/src/modelparams/_generated/catalog.json b/packages/modelparams-python/src/modelparams/_generated/catalog.json index 7e184d1d..590b8f45 100644 --- a/packages/modelparams-python/src/modelparams/_generated/catalog.json +++ b/packages/modelparams-python/src/modelparams/_generated/catalog.json @@ -16516,6 +16516,95 @@ } ] }, + { + "provider": "fireworks", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "minimax-m3p1-flash", + "wireId": "accounts/fireworks/models/minimax-m3p1-flash", + "params": [ + { + "path": "max_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the response.", + "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": "top_k", + "label": "Top K", + "description": "Limits token sampling to the top K most likely next tokens.", + "group": "sampling", + "type": "integer", + "range": { + "min": 1, + "max": 100 + } + }, + { + "path": "presence_penalty", + "label": "Presence penalty", + "description": "Penalizes tokens that have already appeared to encourage a wider variety of content.", + "group": "sampling", + "type": "number", + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "frequency_penalty", + "label": "Frequency penalty", + "description": "Penalizes tokens in proportion to how often they have already appeared.", + "group": "sampling", + "type": "number", + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "response_format.type", + "label": "Response format", + "description": "Choose plain text or force valid JSON output.", + "group": "output_format", + "type": "enum", + "values": [ + "text", + "json_object" + ] + } + ] + }, { "provider": "fireworks", "authType": "api_key", @@ -21275,6 +21364,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21372,6 +21462,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.1", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21424,6 +21515,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.1-highspeed", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21566,6 +21658,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.5", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21618,6 +21711,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.5-highspeed", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21760,6 +21854,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.7", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21812,6 +21907,7 @@ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.7-highspeed", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -25227,6 +25323,103 @@ } ] }, + { + "provider": "mistral", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "zai-glm-5-3-flash", + "params": [ + { + "path": "max_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the completion.", + "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": 1.5, + "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", + "default": 1, + "range": { + "min": 0.01, + "max": 1, + "step": 0.01 + } + }, + { + "path": "random_seed", + "label": "Random seed", + "description": "Seed used for deterministic sampling when reproducible outputs are desired.", + "group": "sampling", + "type": "integer", + "range": { + "min": 0 + } + }, + { + "path": "presence_penalty", + "label": "Presence penalty", + "description": "Penalizes repeated words or phrases to encourage a wider variety of generated content.", + "group": "sampling", + "type": "number", + "default": 0, + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "frequency_penalty", + "label": "Frequency penalty", + "description": "Penalizes words based on how often they already appear in the generated text.", + "group": "sampling", + "type": "number", + "default": 0, + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "prompt_mode", + "label": "Prompt mode", + "description": "Enables Mistral's reasoning system prompt; leave unset to disable the default reasoning behavior.", + "group": "reasoning", + "type": "enum", + "values": [ + "reasoning" + ] + }, + { + "path": "safe_prompt", + "label": "Safe prompt", + "description": "Controls whether Mistral injects its safety prompt before the conversation.", + "group": "provider_metadata", + "type": "boolean", + "default": false + } + ] + }, { "provider": "mistral", "authType": "api_key", @@ -25899,6 +26092,188 @@ } ] }, + { + "provider": "moonshot", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "kimi-k3.1-pro", + "status": "active", + "params": [ + { + "path": "max_completion_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the chat completion.", + "group": "generation_length", + "type": "integer", + "range": { + "min": 1 + } + }, + { + "path": "thinking.type", + "label": "Thinking mode", + "description": "Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.", + "group": "reasoning", + "type": "enum", + "default": "enabled", + "values": [ + "enabled", + "disabled" + ] + }, + { + "path": "thinking.keep", + "label": "Keep prior reasoning", + "description": "Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.", + "group": "reasoning", + "type": "enum", + "default": null, + "values": [ + "all", + null + ] + }, + { + "path": "response_format.type", + "label": "Response format", + "description": "Forces the response into plain text or a JSON object.", + "group": "output_format", + "type": "enum", + "default": "text", + "values": [ + "text", + "json_object" + ] + }, + { + "path": "tool_choice", + "label": "Tool choice", + "description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.", + "group": "tooling", + "type": "enum", + "default": "auto", + "values": [ + "auto", + "none" + ] + }, + { + "path": "logprobs", + "label": "Log probabilities", + "description": "Controls whether the response includes log probabilities for the generated tokens.", + "group": "observability", + "type": "boolean", + "default": false + }, + { + "path": "top_logprobs", + "label": "Top log probabilities", + "description": "Number of most likely tokens to return a log probability for at each position.", + "group": "observability", + "applicability": { + "only": { + "logprobs": true + } + }, + "type": "integer", + "range": { + "min": 0, + "max": 20 + } + } + ] + }, + { + "provider": "moonshot", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "kimi-k3.1-pro-highspeed", + "status": "active", + "params": [ + { + "path": "max_completion_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the chat completion.", + "group": "generation_length", + "type": "integer", + "range": { + "min": 1 + } + }, + { + "path": "thinking.type", + "label": "Thinking mode", + "description": "Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.", + "group": "reasoning", + "type": "enum", + "default": "enabled", + "values": [ + "enabled", + "disabled" + ] + }, + { + "path": "thinking.keep", + "label": "Keep prior reasoning", + "description": "Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.", + "group": "reasoning", + "type": "enum", + "default": null, + "values": [ + "all", + null + ] + }, + { + "path": "response_format.type", + "label": "Response format", + "description": "Forces the response into plain text or a JSON object.", + "group": "output_format", + "type": "enum", + "default": "text", + "values": [ + "text", + "json_object" + ] + }, + { + "path": "tool_choice", + "label": "Tool choice", + "description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.", + "group": "tooling", + "type": "enum", + "default": "auto", + "values": [ + "auto", + "none" + ] + }, + { + "path": "logprobs", + "label": "Log probabilities", + "description": "Controls whether the response includes log probabilities for the generated tokens.", + "group": "observability", + "type": "boolean", + "default": false + }, + { + "path": "top_logprobs", + "label": "Top log probabilities", + "description": "Number of most likely tokens to return a log probability for at each position.", + "group": "observability", + "applicability": { + "only": { + "logprobs": true + } + }, + "type": "integer", + "range": { + "min": 0, + "max": 20 + } + } + ] + }, { "provider": "moonshot", "authType": "api_key", @@ -26970,6 +27345,7 @@ "apiSurface": "openai-chat-completions", "model": "kimi-k3", "wireId": "moonshotai/kimi-k3", + "status": "retired", "params": [ { "path": "temperature", @@ -28307,6 +28683,74 @@ } ] }, + { + "provider": "nvidia", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "nemotron-3.5-bolt-e4b", + "wireId": "nvidia/nemotron-3.5-bolt-e4b", + "params": [ + { + "path": "temperature", + "label": "Temperature", + "description": "Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.", + "group": "sampling", + "type": "number", + "default": 1, + "range": { + "max": 1 + } + }, + { + "path": "top_p", + "label": "Top P", + "description": "Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.", + "group": "sampling", + "type": "number", + "default": 0.95, + "range": { + "max": 1 + } + }, + { + "path": "max_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate. Generation stops when this limit is reached.", + "group": "generation_length", + "type": "integer", + "default": 16384, + "range": { + "min": 1, + "max": 32768 + } + }, + { + "path": "reasoning_effort", + "label": "Reasoning effort", + "description": "Controls the reasoning mode. 'none' disables reasoning tokens, 'low' enables low-effort reasoning, and 'high' enables full reasoning.", + "group": "reasoning", + "type": "enum", + "default": "high", + "values": [ + "none", + "low", + "high" + ] + }, + { + "path": "reasoning_budget", + "label": "Reasoning budget", + "description": "Maximum number of tokens the model may use for internal reasoning before being forced to end the reasoning trace. Use -1 to disable budget enforcement.", + "group": "reasoning", + "type": "integer", + "default": 16384, + "range": { + "min": -1, + "max": 32768 + } + } + ] + }, { "provider": "nvidia", "authType": "api_key", @@ -30479,7 +30923,7 @@ "authType": "subscription", "apiSurface": "openai-responses", "model": "gpt-5.5", - "status": "active", + "status": "retired", "params": [ { "path": "reasoning.effort", diff --git a/packages/modelparams-python/src/modelparams/_generated/model_ids.py b/packages/modelparams-python/src/modelparams/_generated/model_ids.py index 03dc733b..f67df562 100644 --- a/packages/modelparams-python/src/modelparams/_generated/model_ids.py +++ b/packages/modelparams-python/src/modelparams/_generated/model_ids.py @@ -218,6 +218,7 @@ "fireworks/kimi-k3", "fireworks/minimax-m2p7", "fireworks/minimax-m3", + "fireworks/minimax-m3p1-flash", "fireworks/muse-glimmer-30b", "fireworks/muse-spark-open", "fireworks/nemotron-3-ultra-nvfp4", @@ -318,6 +319,7 @@ "mistral/zai-glm-5", "mistral/zai-glm-5-2", "mistral/zai-glm-5-3", + "mistral/zai-glm-5-3-flash", "mistral/zai-glm-latest", "moonshot/kimi-k2.5", "moonshot/kimi-k2.6", @@ -327,6 +329,8 @@ "moonshot/kimi-k2.7-code-highspeed-subscription", "moonshot/kimi-k2.7-code-subscription", "moonshot/kimi-k3", + "moonshot/kimi-k3.1-pro", + "moonshot/kimi-k3.1-pro-highspeed", "moonshot/moonshot-v1-128k", "moonshot/moonshot-v1-32k", "moonshot/moonshot-v1-8k", @@ -357,6 +361,7 @@ "nvidia/nemotron-3-super-120b-a12b", "nvidia/nemotron-3-ultra-550b-a55b", "nvidia/nemotron-3-ultra-subscription", + "nvidia/nemotron-3.5-bolt-e4b", "nvidia/nemotron-3.5-lightning-30b-a3b", "nvidia/nemotron-3.5-super-vl-120b-a12b", "nvidia/nemotron-content-safety-reasoning-4b", @@ -718,6 +723,7 @@ "fireworks/kimi-k3", "fireworks/minimax-m2p7", "fireworks/minimax-m3", + "fireworks/minimax-m3p1-flash", "fireworks/muse-glimmer-30b", "fireworks/muse-spark-open", "fireworks/nemotron-3-ultra-nvfp4", @@ -818,6 +824,7 @@ "mistral/zai-glm-5", "mistral/zai-glm-5-2", "mistral/zai-glm-5-3", + "mistral/zai-glm-5-3-flash", "mistral/zai-glm-latest", "moonshot/kimi-k2.5", "moonshot/kimi-k2.6", @@ -827,6 +834,8 @@ "moonshot/kimi-k2.7-code-highspeed-subscription", "moonshot/kimi-k2.7-code-subscription", "moonshot/kimi-k3", + "moonshot/kimi-k3.1-pro", + "moonshot/kimi-k3.1-pro-highspeed", "moonshot/moonshot-v1-128k", "moonshot/moonshot-v1-32k", "moonshot/moonshot-v1-8k", @@ -857,6 +866,7 @@ "nvidia/nemotron-3-super-120b-a12b", "nvidia/nemotron-3-ultra-550b-a55b", "nvidia/nemotron-3-ultra-subscription", + "nvidia/nemotron-3.5-bolt-e4b", "nvidia/nemotron-3.5-lightning-30b-a3b", "nvidia/nemotron-3.5-super-vl-120b-a12b", "nvidia/nemotron-content-safety-reasoning-4b", diff --git a/packages/modelparams-python/src/modelparams/_generated/registry.py b/packages/modelparams-python/src/modelparams/_generated/registry.py index c4b3900f..d44296ad 100644 --- a/packages/modelparams-python/src/modelparams/_generated/registry.py +++ b/packages/modelparams-python/src/modelparams/_generated/registry.py @@ -243,6 +243,7 @@ "fireworks/kimi-k3": fireworks.Kimi_K3Params, "fireworks/minimax-m2p7": fireworks.Minimax_M2p7Params, "fireworks/minimax-m3": fireworks.Minimax_M3Params, + "fireworks/minimax-m3p1-flash": fireworks.Minimax_M3p1_FlashParams, "fireworks/muse-glimmer-30b": fireworks.Muse_Glimmer_30bParams, "fireworks/muse-spark-open": fireworks.Muse_Spark_OpenParams, "fireworks/nemotron-3-ultra-nvfp4": fireworks.Nemotron_3_Ultra_Nvfp4Params, @@ -343,6 +344,7 @@ "mistral/zai-glm-5": mistral.Zai_Glm_5Params, "mistral/zai-glm-5-2": mistral.Zai_Glm_5_2Params, "mistral/zai-glm-5-3": mistral.Zai_Glm_5_3Params, + "mistral/zai-glm-5-3-flash": mistral.Zai_Glm_5_3_FlashParams, "mistral/zai-glm-latest": mistral.Zai_Glm_LatestParams, "moonshot/kimi-k2.5": moonshot.Kimi_K2_5Params, "moonshot/kimi-k2.6": moonshot.Kimi_K2_6Params, @@ -352,6 +354,8 @@ "moonshot/kimi-k2.7-code-highspeed-subscription": moonshot.Kimi_K2_7_Code_Highspeed_SubscriptionParams, "moonshot/kimi-k2.7-code-subscription": moonshot.Kimi_K2_7_Code_SubscriptionParams, "moonshot/kimi-k3": moonshot.Kimi_K3Params, + "moonshot/kimi-k3.1-pro": moonshot.Kimi_K3_1_ProParams, + "moonshot/kimi-k3.1-pro-highspeed": moonshot.Kimi_K3_1_Pro_HighspeedParams, "moonshot/moonshot-v1-128k": moonshot.Moonshot_V1_128kParams, "moonshot/moonshot-v1-32k": moonshot.Moonshot_V1_32kParams, "moonshot/moonshot-v1-8k": moonshot.Moonshot_V1_8kParams, @@ -382,6 +386,7 @@ "nvidia/nemotron-3-super-120b-a12b": nvidia.Nemotron_3_Super_120b_A12bParams, "nvidia/nemotron-3-ultra-550b-a55b": nvidia.Nemotron_3_Ultra_550b_A55bParams, "nvidia/nemotron-3-ultra-subscription": nvidia.Nemotron_3_Ultra_SubscriptionParams, + "nvidia/nemotron-3.5-bolt-e4b": nvidia.Nemotron_3_5_Bolt_E4bParams, "nvidia/nemotron-3.5-lightning-30b-a3b": nvidia.Nemotron_3_5_Lightning_30b_A3bParams, "nvidia/nemotron-3.5-super-vl-120b-a12b": nvidia.Nemotron_3_5_Super_Vl_120b_A12bParams, "nvidia/nemotron-content-safety-reasoning-4b": nvidia.Nemotron_Content_Safety_Reasoning_4bParams, diff --git a/packages/modelparams-python/src/modelparams/types/fireworks.py b/packages/modelparams-python/src/modelparams/types/fireworks.py index ac70a28f..df72ee75 100644 --- a/packages/modelparams-python/src/modelparams/types/fireworks.py +++ b/packages/modelparams-python/src/modelparams/types/fireworks.py @@ -280,6 +280,21 @@ ) setattr(Minimax_M3Params, "__pydantic_config__", _PARAMS_CONFIG) +Minimax_M3p1_FlashParams = TypedDict( + "Minimax_M3p1_FlashParams", + { + "max_tokens": Annotated[int, Field(ge=1)], + "temperature": Annotated[float, Field(ge=0, le=2)], + "top_p": Annotated[float, Field(ge=0, le=1)], + "top_k": Annotated[int, Field(ge=1, le=100)], + "presence_penalty": Annotated[float, Field(ge=-2, le=2)], + "frequency_penalty": Annotated[float, Field(ge=-2, le=2)], + "response_format.type": Literal["text", "json_object"], + }, + total=False, +) +setattr(Minimax_M3p1_FlashParams, "__pydantic_config__", _PARAMS_CONFIG) + Muse_Glimmer_30bParams = TypedDict( "Muse_Glimmer_30bParams", { @@ -404,6 +419,7 @@ "Kimi_K3Params", "Minimax_M2p7Params", "Minimax_M3Params", + "Minimax_M3p1_FlashParams", "Muse_Glimmer_30bParams", "Muse_Spark_OpenParams", "Nemotron_3_Ultra_Nvfp4Params", diff --git a/packages/modelparams-python/src/modelparams/types/mistral.py b/packages/modelparams-python/src/modelparams/types/mistral.py index 3f850158..0ba19b34 100644 --- a/packages/modelparams-python/src/modelparams/types/mistral.py +++ b/packages/modelparams-python/src/modelparams/types/mistral.py @@ -523,6 +523,22 @@ ) setattr(Zai_Glm_5_3Params, "__pydantic_config__", _PARAMS_CONFIG) +Zai_Glm_5_3_FlashParams = TypedDict( + "Zai_Glm_5_3_FlashParams", + { + "max_tokens": Annotated[int, Field(ge=1)], + "temperature": Annotated[float, Field(ge=0, le=1.5)], + "top_p": Annotated[float, Field(ge=0.01, le=1)], + "random_seed": Annotated[int, Field(ge=0)], + "presence_penalty": Annotated[float, Field(ge=-2, le=2)], + "frequency_penalty": Annotated[float, Field(ge=-2, le=2)], + "prompt_mode": Literal["reasoning"], + "safe_prompt": bool, + }, + total=False, +) +setattr(Zai_Glm_5_3_FlashParams, "__pydantic_config__", _PARAMS_CONFIG) + Zai_Glm_LatestParams = TypedDict( "Zai_Glm_LatestParams", { @@ -571,5 +587,6 @@ "Zai_Glm_5Params", "Zai_Glm_5_2Params", "Zai_Glm_5_3Params", + "Zai_Glm_5_3_FlashParams", "Zai_Glm_LatestParams", ] diff --git a/packages/modelparams-python/src/modelparams/types/moonshot.py b/packages/modelparams-python/src/modelparams/types/moonshot.py index ef160b00..311cef07 100644 --- a/packages/modelparams-python/src/modelparams/types/moonshot.py +++ b/packages/modelparams-python/src/modelparams/types/moonshot.py @@ -119,6 +119,36 @@ ) setattr(Kimi_K3Params, "__pydantic_config__", _PARAMS_CONFIG) +Kimi_K3_1_ProParams = TypedDict( + "Kimi_K3_1_ProParams", + { + "max_completion_tokens": Annotated[int, Field(ge=1)], + "thinking.type": Literal["enabled", "disabled"], + "thinking.keep": Literal["all", None], + "response_format.type": Literal["text", "json_object"], + "tool_choice": Literal["auto", "none"], + "logprobs": bool, + "top_logprobs": Annotated[int, Field(ge=0, le=20)], + }, + total=False, +) +setattr(Kimi_K3_1_ProParams, "__pydantic_config__", _PARAMS_CONFIG) + +Kimi_K3_1_Pro_HighspeedParams = TypedDict( + "Kimi_K3_1_Pro_HighspeedParams", + { + "max_completion_tokens": Annotated[int, Field(ge=1)], + "thinking.type": Literal["enabled", "disabled"], + "thinking.keep": Literal["all", None], + "response_format.type": Literal["text", "json_object"], + "tool_choice": Literal["auto", "none"], + "logprobs": bool, + "top_logprobs": Annotated[int, Field(ge=0, le=20)], + }, + total=False, +) +setattr(Kimi_K3_1_Pro_HighspeedParams, "__pydantic_config__", _PARAMS_CONFIG) + Moonshot_V1_128kParams = TypedDict( "Moonshot_V1_128kParams", { @@ -185,6 +215,8 @@ "Kimi_K2_7_Code_Highspeed_SubscriptionParams", "Kimi_K2_7_Code_SubscriptionParams", "Kimi_K3Params", + "Kimi_K3_1_ProParams", + "Kimi_K3_1_Pro_HighspeedParams", "Moonshot_V1_128kParams", "Moonshot_V1_32kParams", "Moonshot_V1_8kParams", diff --git a/packages/modelparams-python/src/modelparams/types/nvidia.py b/packages/modelparams-python/src/modelparams/types/nvidia.py index c0ec3c74..604b807a 100644 --- a/packages/modelparams-python/src/modelparams/types/nvidia.py +++ b/packages/modelparams-python/src/modelparams/types/nvidia.py @@ -384,6 +384,19 @@ ) setattr(Nemotron_3_Ultra_SubscriptionParams, "__pydantic_config__", _PARAMS_CONFIG) +Nemotron_3_5_Bolt_E4bParams = TypedDict( + "Nemotron_3_5_Bolt_E4bParams", + { + "temperature": Annotated[float, Field(le=1)], + "top_p": Annotated[float, Field(le=1)], + "max_tokens": Annotated[int, Field(ge=1, le=32768)], + "reasoning_effort": Literal["none", "low", "high"], + "reasoning_budget": Annotated[int, Field(ge=-1, le=32768)], + }, + total=False, +) +setattr(Nemotron_3_5_Bolt_E4bParams, "__pydantic_config__", _PARAMS_CONFIG) + Nemotron_3_5_Lightning_30b_A3bParams = TypedDict( "Nemotron_3_5_Lightning_30b_A3bParams", { @@ -504,6 +517,7 @@ "Nemotron_3_Super_120b_A12bParams", "Nemotron_3_Ultra_550b_A55bParams", "Nemotron_3_Ultra_SubscriptionParams", + "Nemotron_3_5_Bolt_E4bParams", "Nemotron_3_5_Lightning_30b_A3bParams", "Nemotron_3_5_Super_Vl_120b_A12bParams", "Nemotron_Content_Safety_Reasoning_4bParams", diff --git a/packages/modelparams/src/generated/data.ts b/packages/modelparams/src/generated/data.ts index c978a8cb..3d1dd2cc 100644 --- a/packages/modelparams/src/generated/data.ts +++ b/packages/modelparams/src/generated/data.ts @@ -16521,6 +16521,95 @@ const GENERATED_CATALOG = [ } ] }, + { + "provider": "fireworks", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "minimax-m3p1-flash", + "wireId": "accounts/fireworks/models/minimax-m3p1-flash", + "params": [ + { + "path": "max_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the response.", + "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": "top_k", + "label": "Top K", + "description": "Limits token sampling to the top K most likely next tokens.", + "group": "sampling", + "type": "integer", + "range": { + "min": 1, + "max": 100 + } + }, + { + "path": "presence_penalty", + "label": "Presence penalty", + "description": "Penalizes tokens that have already appeared to encourage a wider variety of content.", + "group": "sampling", + "type": "number", + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "frequency_penalty", + "label": "Frequency penalty", + "description": "Penalizes tokens in proportion to how often they have already appeared.", + "group": "sampling", + "type": "number", + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "response_format.type", + "label": "Response format", + "description": "Choose plain text or force valid JSON output.", + "group": "output_format", + "type": "enum", + "values": [ + "text", + "json_object" + ] + } + ] + }, { "provider": "fireworks", "authType": "api_key", @@ -21280,6 +21369,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21377,6 +21467,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.1", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21429,6 +21520,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.1-highspeed", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21571,6 +21663,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.5", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21623,6 +21716,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.5-highspeed", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21765,6 +21859,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.7", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -21817,6 +21912,7 @@ const GENERATED_CATALOG = [ "authType": "api_key", "apiSurface": "openai-chat-completions", "model": "MiniMax-M2.7-highspeed", + "status": "retired", "params": [ { "path": "max_completion_tokens", @@ -25232,6 +25328,103 @@ const GENERATED_CATALOG = [ } ] }, + { + "provider": "mistral", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "zai-glm-5-3-flash", + "params": [ + { + "path": "max_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the completion.", + "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": 1.5, + "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", + "default": 1, + "range": { + "min": 0.01, + "max": 1, + "step": 0.01 + } + }, + { + "path": "random_seed", + "label": "Random seed", + "description": "Seed used for deterministic sampling when reproducible outputs are desired.", + "group": "sampling", + "type": "integer", + "range": { + "min": 0 + } + }, + { + "path": "presence_penalty", + "label": "Presence penalty", + "description": "Penalizes repeated words or phrases to encourage a wider variety of generated content.", + "group": "sampling", + "type": "number", + "default": 0, + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "frequency_penalty", + "label": "Frequency penalty", + "description": "Penalizes words based on how often they already appear in the generated text.", + "group": "sampling", + "type": "number", + "default": 0, + "range": { + "min": -2, + "max": 2, + "step": 0.1 + } + }, + { + "path": "prompt_mode", + "label": "Prompt mode", + "description": "Enables Mistral's reasoning system prompt; leave unset to disable the default reasoning behavior.", + "group": "reasoning", + "type": "enum", + "values": [ + "reasoning" + ] + }, + { + "path": "safe_prompt", + "label": "Safe prompt", + "description": "Controls whether Mistral injects its safety prompt before the conversation.", + "group": "provider_metadata", + "type": "boolean", + "default": false + } + ] + }, { "provider": "mistral", "authType": "api_key", @@ -25904,6 +26097,188 @@ const GENERATED_CATALOG = [ } ] }, + { + "provider": "moonshot", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "kimi-k3.1-pro", + "status": "active", + "params": [ + { + "path": "max_completion_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the chat completion.", + "group": "generation_length", + "type": "integer", + "range": { + "min": 1 + } + }, + { + "path": "thinking.type", + "label": "Thinking mode", + "description": "Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.", + "group": "reasoning", + "type": "enum", + "default": "enabled", + "values": [ + "enabled", + "disabled" + ] + }, + { + "path": "thinking.keep", + "label": "Keep prior reasoning", + "description": "Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.", + "group": "reasoning", + "type": "enum", + "default": null, + "values": [ + "all", + null + ] + }, + { + "path": "response_format.type", + "label": "Response format", + "description": "Forces the response into plain text or a JSON object.", + "group": "output_format", + "type": "enum", + "default": "text", + "values": [ + "text", + "json_object" + ] + }, + { + "path": "tool_choice", + "label": "Tool choice", + "description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.", + "group": "tooling", + "type": "enum", + "default": "auto", + "values": [ + "auto", + "none" + ] + }, + { + "path": "logprobs", + "label": "Log probabilities", + "description": "Controls whether the response includes log probabilities for the generated tokens.", + "group": "observability", + "type": "boolean", + "default": false + }, + { + "path": "top_logprobs", + "label": "Top log probabilities", + "description": "Number of most likely tokens to return a log probability for at each position.", + "group": "observability", + "applicability": { + "only": { + "logprobs": true + } + }, + "type": "integer", + "range": { + "min": 0, + "max": 20 + } + } + ] + }, + { + "provider": "moonshot", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "kimi-k3.1-pro-highspeed", + "status": "active", + "params": [ + { + "path": "max_completion_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate in the chat completion.", + "group": "generation_length", + "type": "integer", + "range": { + "min": 1 + } + }, + { + "path": "thinking.type", + "label": "Thinking mode", + "description": "Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.", + "group": "reasoning", + "type": "enum", + "default": "enabled", + "values": [ + "enabled", + "disabled" + ] + }, + { + "path": "thinking.keep", + "label": "Keep prior reasoning", + "description": "Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.", + "group": "reasoning", + "type": "enum", + "default": null, + "values": [ + "all", + null + ] + }, + { + "path": "response_format.type", + "label": "Response format", + "description": "Forces the response into plain text or a JSON object.", + "group": "output_format", + "type": "enum", + "default": "text", + "values": [ + "text", + "json_object" + ] + }, + { + "path": "tool_choice", + "label": "Tool choice", + "description": "Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.", + "group": "tooling", + "type": "enum", + "default": "auto", + "values": [ + "auto", + "none" + ] + }, + { + "path": "logprobs", + "label": "Log probabilities", + "description": "Controls whether the response includes log probabilities for the generated tokens.", + "group": "observability", + "type": "boolean", + "default": false + }, + { + "path": "top_logprobs", + "label": "Top log probabilities", + "description": "Number of most likely tokens to return a log probability for at each position.", + "group": "observability", + "applicability": { + "only": { + "logprobs": true + } + }, + "type": "integer", + "range": { + "min": 0, + "max": 20 + } + } + ] + }, { "provider": "moonshot", "authType": "api_key", @@ -26975,6 +27350,7 @@ const GENERATED_CATALOG = [ "apiSurface": "openai-chat-completions", "model": "kimi-k3", "wireId": "moonshotai/kimi-k3", + "status": "retired", "params": [ { "path": "temperature", @@ -28312,6 +28688,74 @@ const GENERATED_CATALOG = [ } ] }, + { + "provider": "nvidia", + "authType": "api_key", + "apiSurface": "openai-chat-completions", + "model": "nemotron-3.5-bolt-e4b", + "wireId": "nvidia/nemotron-3.5-bolt-e4b", + "params": [ + { + "path": "temperature", + "label": "Temperature", + "description": "Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.", + "group": "sampling", + "type": "number", + "default": 1, + "range": { + "max": 1 + } + }, + { + "path": "top_p", + "label": "Top P", + "description": "Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.", + "group": "sampling", + "type": "number", + "default": 0.95, + "range": { + "max": 1 + } + }, + { + "path": "max_tokens", + "label": "Max tokens", + "description": "Maximum number of tokens to generate. Generation stops when this limit is reached.", + "group": "generation_length", + "type": "integer", + "default": 16384, + "range": { + "min": 1, + "max": 32768 + } + }, + { + "path": "reasoning_effort", + "label": "Reasoning effort", + "description": "Controls the reasoning mode. 'none' disables reasoning tokens, 'low' enables low-effort reasoning, and 'high' enables full reasoning.", + "group": "reasoning", + "type": "enum", + "default": "high", + "values": [ + "none", + "low", + "high" + ] + }, + { + "path": "reasoning_budget", + "label": "Reasoning budget", + "description": "Maximum number of tokens the model may use for internal reasoning before being forced to end the reasoning trace. Use -1 to disable budget enforcement.", + "group": "reasoning", + "type": "integer", + "default": 16384, + "range": { + "min": -1, + "max": 32768 + } + } + ] + }, { "provider": "nvidia", "authType": "api_key", @@ -30484,7 +30928,7 @@ const GENERATED_CATALOG = [ "authType": "subscription", "apiSurface": "openai-responses", "model": "gpt-5.5", - "status": "active", + "status": "retired", "params": [ { "path": "reasoning.effort", diff --git a/packages/modelparams/src/generated/defaults.ts b/packages/modelparams/src/generated/defaults.ts index 9d7493c2..220cdafd 100644 --- a/packages/modelparams/src/generated/defaults.ts +++ b/packages/modelparams/src/generated/defaults.ts @@ -865,6 +865,7 @@ export const DEFAULTS = { "fireworks/kimi-k3": {}, "fireworks/minimax-m2p7": {}, "fireworks/minimax-m3": {}, + "fireworks/minimax-m3p1-flash": {}, "fireworks/muse-glimmer-30b": {}, "fireworks/muse-spark-open": {}, "fireworks/nemotron-3-ultra-nvfp4": {}, @@ -1468,6 +1469,12 @@ export const DEFAULTS = { frequency_penalty: 0, safe_prompt: false, }, + "mistral/zai-glm-5-3-flash": { + top_p: 1, + presence_penalty: 0, + frequency_penalty: 0, + safe_prompt: false, + }, "mistral/zai-glm-latest": { top_p: 1, presence_penalty: 0, @@ -1519,6 +1526,20 @@ export const DEFAULTS = { "response_format.type": "text", tool_choice: "auto", }, + "moonshot/kimi-k3.1-pro": { + "thinking.type": "enabled", + "thinking.keep": null, + "response_format.type": "text", + tool_choice: "auto", + logprobs: false, + }, + "moonshot/kimi-k3.1-pro-highspeed": { + "thinking.type": "enabled", + "thinking.keep": null, + "response_format.type": "text", + tool_choice: "auto", + logprobs: false, + }, "moonshot/moonshot-v1-128k": { temperature: 0.3, top_p: 1, @@ -1737,6 +1758,13 @@ export const DEFAULTS = { reasoning_effort: "high", reasoning_budget: 16384, }, + "nvidia/nemotron-3.5-bolt-e4b": { + temperature: 1, + top_p: 0.95, + max_tokens: 16384, + reasoning_effort: "high", + reasoning_budget: 16384, + }, "nvidia/nemotron-3.5-lightning-30b-a3b": { temperature: 1, top_p: 0.95, diff --git a/packages/modelparams/src/generated/model-ids.ts b/packages/modelparams/src/generated/model-ids.ts index 413e6e4d..114e0019 100644 --- a/packages/modelparams/src/generated/model-ids.ts +++ b/packages/modelparams/src/generated/model-ids.ts @@ -216,6 +216,7 @@ export const MODEL_IDS = [ "fireworks/kimi-k3", "fireworks/minimax-m2p7", "fireworks/minimax-m3", + "fireworks/minimax-m3p1-flash", "fireworks/muse-glimmer-30b", "fireworks/muse-spark-open", "fireworks/nemotron-3-ultra-nvfp4", @@ -316,6 +317,7 @@ export const MODEL_IDS = [ "mistral/zai-glm-5", "mistral/zai-glm-5-2", "mistral/zai-glm-5-3", + "mistral/zai-glm-5-3-flash", "mistral/zai-glm-latest", "moonshot/kimi-k2.5", "moonshot/kimi-k2.6", @@ -325,6 +327,8 @@ export const MODEL_IDS = [ "moonshot/kimi-k2.7-code-highspeed-subscription", "moonshot/kimi-k2.7-code-subscription", "moonshot/kimi-k3", + "moonshot/kimi-k3.1-pro", + "moonshot/kimi-k3.1-pro-highspeed", "moonshot/moonshot-v1-128k", "moonshot/moonshot-v1-32k", "moonshot/moonshot-v1-8k", @@ -355,6 +359,7 @@ export const MODEL_IDS = [ "nvidia/nemotron-3-super-120b-a12b", "nvidia/nemotron-3-ultra-550b-a55b", "nvidia/nemotron-3-ultra-subscription", + "nvidia/nemotron-3.5-bolt-e4b", "nvidia/nemotron-3.5-lightning-30b-a3b", "nvidia/nemotron-3.5-super-vl-120b-a12b", "nvidia/nemotron-content-safety-reasoning-4b", diff --git a/packages/modelparams/src/generated/params-by-id.ts b/packages/modelparams/src/generated/params-by-id.ts index 3ab70ea6..64afbb91 100644 --- a/packages/modelparams/src/generated/params-by-id.ts +++ b/packages/modelparams/src/generated/params-by-id.ts @@ -1620,6 +1620,15 @@ export type ParamsById = { frequency_penalty: number; "response_format.type": "text" | "json_object"; }; + "fireworks/minimax-m3p1-flash": { + max_tokens: number; + temperature: number; + top_p: number; + top_k: number; + presence_penalty: number; + frequency_penalty: number; + "response_format.type": "text" | "json_object"; + }; "fireworks/muse-glimmer-30b": { max_tokens: number; temperature: number; @@ -2532,6 +2541,16 @@ export type ParamsById = { prompt_mode: "reasoning"; safe_prompt: boolean; }; + "mistral/zai-glm-5-3-flash": { + max_tokens: number; + temperature: number; + top_p: number; + random_seed: number; + presence_penalty: number; + frequency_penalty: number; + prompt_mode: "reasoning"; + safe_prompt: boolean; + }; "mistral/zai-glm-latest": { max_tokens: number; temperature: number; @@ -2603,6 +2622,24 @@ export type ParamsById = { "response_format.type": "text" | "json_object" | "json_schema"; tool_choice: "auto" | "none" | "required"; }; + "moonshot/kimi-k3.1-pro": { + max_completion_tokens: number; + "thinking.type": "enabled" | "disabled"; + "thinking.keep": "all" | null; + "response_format.type": "text" | "json_object"; + tool_choice: "auto" | "none"; + logprobs: boolean; + top_logprobs: number; + }; + "moonshot/kimi-k3.1-pro-highspeed": { + max_completion_tokens: number; + "thinking.type": "enabled" | "disabled"; + "thinking.keep": "all" | null; + "response_format.type": "text" | "json_object"; + tool_choice: "auto" | "none"; + logprobs: boolean; + top_logprobs: number; + }; "moonshot/moonshot-v1-128k": { max_completion_tokens: number; stop: string; @@ -2852,6 +2889,13 @@ export type ParamsById = { reasoning_budget: number; stop: string; }; + "nvidia/nemotron-3.5-bolt-e4b": { + temperature: number; + top_p: number; + max_tokens: number; + reasoning_effort: "none" | "low" | "high"; + reasoning_budget: number; + }; "nvidia/nemotron-3.5-lightning-30b-a3b": { temperature: number; top_p: number;