unique_toolkit 2026.40.0.dev5__py3-none-any.whl → 2026.40.0.dev7__py3-none-any.whl

unique_toolkit/language_model/infos.py

@@ -113,6 +113,7 @@ class LanguageModelName(StrEnum):
     GROK_4_1_FAST_REASONING = "litellm:grok-4-1-fast-reasoning"
     GROK_4_5 = "litellm:grok-4-5"
     GROK_4_6 = "litellm:grok-4-6"
+    GROK_4_7 = "litellm:grok-4-7"
     LITELLM_OPENAI_GPT_5 = "litellm:openai-gpt-5"
     LITELLM_OPENAI_GPT_5_MINI = "litellm:openai-gpt-5-mini"
     LITELLM_OPENAI_GPT_5_NANO = "litellm:openai-gpt-5-nano"
@@ -141,6 +142,7 @@ class LanguageModelName(StrEnum):
     LITELLM_DEEPSEEK_R1 = "litellm:deepseek-r1"
     LITELLM_DEEPSEEK_V3 = "litellm:deepseek-v3-1"
     LITELLM_DEEPSEEK_V4_PRO = "litellm:deepseek-v4-pro"
+    LITELLM_DEEPSEEK_V4_1_FLASH = "litellm:deepseek-v4-1-flash"
     LITELLM_GLM_5_1 = "litellm:glm-5.1"
     LITELLM_GLM_5_2 = "litellm:glm-5.2"
     LITELLM_PHOENIQS_DEEPSEEK_V4_FLASH = "litellm:phoeniqs-deepseek-v4-flash"
@@ -2322,6 +2324,34 @@ class LanguageModelInfo(BaseModel):
                     supported_reasoning_efforts=["low", "medium", "high", "xhigh"],
                     default_options={"reasoning_effort": "high"},
                 )
+            case LanguageModelName.GROK_4_7:
+                # Specs from https://docs.x.ai/developers/models/grok-4.7
+                # Knowledge cutoff from https://docs.x.ai/developers/models
+                # Published 2026-09-21: https://vercel.com/changelog/grok-4-7-now-available-and-40-off-on-ai-gateway-fx-eve
+                return cls(
+                    name=model_name,
+                    capabilities=[
+                        ModelCapabilities.FUNCTION_CALLING,
+                        ModelCapabilities.REASONING,
+                        ModelCapabilities.STREAMING,
+                        ModelCapabilities.STRUCTURED_OUTPUT,
+                        ModelCapabilities.VISION,
+                    ],
+                    provider=LanguageModelProvider.LITELLM,
+                    family=ModelFamily.XAI,
+                    version="grok-4.7",
+                    encoder_name=EncoderName.O200K_BASE,  # TODO: Update encoder with grok tokenizer
+                    token_limits=LanguageModelTokenLimits(
+                        # Context window is 500_000, we leave 20_000 tokens as buffer due to tokenizer mismatch
+                        # Assign 90% for input and 10% for output
+                        token_limit_input=int(480_000 * 0.9),
+                        token_limit_output=int(480_000 * 0.1),
+                    ),
+                    info_cutoff_at=date(2026, 5, day=1),
+                    published_at=date(2026, 9, day=21),
+                    supported_reasoning_efforts=["low", "medium", "high", "xhigh"],
+                    default_options={"reasoning_effort": "high"},
+                )
             case LanguageModelName.LITELLM_OPENAI_GPT_5:
                 return cls(
                     name=model_name,
@@ -3042,6 +3072,30 @@ class LanguageModelInfo(BaseModel):
                     published_at=date(2026, 1, 1),
                     supported_reasoning_efforts=[],
                 )
+            case LanguageModelName.LITELLM_DEEPSEEK_V4_1_FLASH:
+                return cls(
+                    name=model_name,
+                    provider=LanguageModelProvider.LITELLM,
+                    family=ModelFamily.DEEPSEEK,
+                    version="deepseek-v4-1-flash",
+                    encoder_name=EncoderName.DEEPSEEK,
+                    capabilities=[
+                        ModelCapabilities.FUNCTION_CALLING,
+                        ModelCapabilities.STRUCTURED_OUTPUT,
+                        ModelCapabilities.REASONING,
+                        ModelCapabilities.STREAMING,
+                        ModelCapabilities.VISION,
+                    ],
+                    # 1M context and 384K max output. Alibaba Model Studio lists
+                    # max input 1_000_000 and max output 393_216; Together serves
+                    # maxTokens 384_000.
+                    # https://www.alibabacloud.com/help/en/model-studio/deepseek-v4-1-flash
+                    # https://pi.dev/models/together/deepseek-ai-deepseek-v4-1-flash
+                    token_limits=LanguageModelTokenLimits(
+                        token_limit_input=1_000_000, token_limit_output=384_000
+                    ),
+                    supported_reasoning_efforts=[],
+                )
             case LanguageModelName.LITELLM_GLM_5_1:
                 return cls(
                     name=model_name,

{unique_toolkit-2026.40.0.dev5.dist-info → unique_toolkit-2026.40.0.dev7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: unique-toolkit
-Version: 2026.40.0.dev5
+Version: 2026.40.0.dev7
 Summary: 
 Author: Cedric Klinkert, Martin Fadler, Sadique Sheik, Fabian Schläpfer, Pascal Hauri, Ejiroghene Oki-Peter, Rami Azouz, Ricardo Visini, Sanzio Monti
 Author-email: Cedric Klinkert <cedric.klinkert@unique.ch>, Martin Fadler <martin.fadler@unique.ch>, Sadique Sheik <sadique@unique.ch>, Fabian Schläpfer <fabian@unique.ch>, Pascal Hauri <pascal@unique.ch>, Ejiroghene Oki-Peter <oki-peter.ejiroghene@unique.ai>, Rami Azouz <rami.ext@unique.ch>, Ricardo Visini <ricardo.visini@unique.ch>, Sanzio Monti <sanzio.monti@unique.ai>
@@ -16,7 +16,7 @@ Requires-Dist: sseclient>=0.0.27,<0.0.28
 Requires-Dist: openai>=1.105.0,<3
 Requires-Dist: platformdirs>=4.0.0,<5
 Requires-Dist: pillow>=12.1.1
-Requires-Dist: unique-sdk>=2026.40.0.dev3,<2026.40.0rc0
+Requires-Dist: unique-sdk>=2026.40.0.dev4,<2026.40.0rc0
 Requires-Dist: jambo>=0.1.2,<0.2
 Requires-Dist: jinja2>=3.1.6,<4
 Requires-Dist: docxtpl>=0.20.1,<0.21

{unique_toolkit-2026.40.0.dev5.dist-info → unique_toolkit-2026.40.0.dev7.dist-info}/RECORD

@@ -394,7 +394,7 @@ unique_toolkit/language_model/builder.py,sha256=KwBzsD6-fvQnB4p2I7M-5Tk6j4tmQ973
 unique_toolkit/language_model/constants.py,sha256=B-topqW0r83dkC_25DeQfnPk3n53qzIHUCBS7YJ0-1U,119
 unique_toolkit/language_model/default_language_model.py,sha256=mcyuFs_V6tp52C8xJSHiEHO4kFGj6xed1gvozvCRD7k,1181
 unique_toolkit/language_model/functions.py,sha256=W6JC4LUh-LZ9waDgph--xVmKeQ7Ls6NRAqDx9DwQdeY,26691
-unique_toolkit/language_model/infos.py,sha256=_BGQVZJ9_3DDx4qkh8ejszHW8P6QxDEjJ7X6TIYq9jY,157224
+unique_toolkit/language_model/infos.py,sha256=XXXiDl1Hc_Ew2FsclDhIXQRAAhrkQTnaXoEMlrt5DQk,160244
 unique_toolkit/language_model/invocation_stats.py,sha256=CQbpUz394IsVd8eNTB1gPSMt5f0UXmkN1amNsqtAW6U,5857
 unique_toolkit/language_model/model_costs.py,sha256=ShuDAzMew1fUkJ3tzPgU5Zqjp6ivUZaG97BFh93tEn4,6358
 unique_toolkit/language_model/prompt.py,sha256=JSawaLjQg3VR-E2fK8engFyJnNdk21zaO8pPIodzN4Q,3991
@@ -423,6 +423,6 @@ unique_toolkit/short_term_memory/service.py,sha256=2JMBEBn6GoyK-a0xvdcWpPft-3C74
 unique_toolkit/smart_rules/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 unique_toolkit/smart_rules/compile.py,sha256=vRe0aMJWuVIK-VsqomgxQImzke7MDmwgSxxJfYCa7zw,1386
 unique_toolkit/test_utilities/events.py,sha256=FLrhMh0A_mQutAUjoD0FdMcVQ4Gq31r35OVGzRjvLro,6760
-unique_toolkit-2026.40.0.dev5.dist-info/WHEEL,sha256=eh7sammvW2TypMMMGKgsM83HyA_3qQ5Lgg3ynoecH3M,79
-unique_toolkit-2026.40.0.dev5.dist-info/METADATA,sha256=TB781mnDzgeSR6SyyT5eQZym_pvENv44MlVx-Q8ano4,7821
-unique_toolkit-2026.40.0.dev5.dist-info/RECORD,,
+unique_toolkit-2026.40.0.dev7.dist-info/WHEEL,sha256=eh7sammvW2TypMMMGKgsM83HyA_3qQ5Lgg3ynoecH3M,79
+unique_toolkit-2026.40.0.dev7.dist-info/METADATA,sha256=phNHiRxVXjTPnlFfMP_5jWnpdsYeqvaV2za91B4lzlE,7821
+unique_toolkit-2026.40.0.dev7.dist-info/RECORD,,