python-fastllm 0.0.44__tar.gz → 0.0.46__tar.gz

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-fastllm
-Version: 0.0.44
+Version: 0.0.46
 Author-email: Kerem Turgutlu <keremturgutlu@gmail.com>
 License: Apache-2.0
 Project-URL: Repository, https://github.com/AnswerDotAI/fastllm
@@ -11,7 +11,6 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 Requires-Dist: fastcore>=2.1.18
 Requires-Dist: aidialog>=0.0.10
-Requires-Dist: fastspec>=0.2.1
 Requires-Dist: fasttransport>=0.0.2
 Requires-Dist: pillow
 

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/__init__.py

@@ -2,4 +2,4 @@
 
 - `fastllm.responses`: Provider-independent Responses resources and streaming over FastLLM"""
 
-__version__ = "0.0.44"
+__version__ = "0.0.46"

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/acomplete.py

@@ -10,7 +10,7 @@ import asyncio,json,httpx2
 from fastcore.utils import *
 from fastcore.meta import *
 from fasttransport.core import AsyncHttpCli
-from fastspec.errors import APIError
+from fasttransport.errors import APIError
 
 from .types import *
 from aidialog.msg_parts import (Part, PartType, Msg, Text, Thinking, ToolUse, ToolResult, InputImage, InputAudio, InputVideo,

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/anthropic.py

@@ -12,7 +12,7 @@ import json
 from collections import Counter
 from fastcore.utils import *
 from fastcore.meta import *
-from fastspec.errors import api_error_from_event
+from fasttransport.errors import api_error_from_event
 
 from .types import *
 from aidialog.msg_parts import (Part, PartType, mk_part, Msg, Text, Thinking, ToolUse, ToolResult, tool_text, ServerToolResult,

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/chat.py

@@ -256,7 +256,7 @@ def _prep_call(self:AsyncChat, search, max_tokens, kwargs, stream=False, think=N
     if self.api_key:       kwargs['api_key'] = self.api_key
     if self.base_url:      kwargs['base_url'] = self.base_url
     if self.endpoint:      kwargs['endpoint'] = self.endpoint
-    if self.extra_headers: kwargs['xtra_headers'] = self.extra_headers
+    if self.extra_headers: kwargs['xtra_hdrs'] = self.extra_headers
     kwargs.update(_think_kw(self.model, think, self.vendor_name))
     return max_tokens
 

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/gemini.py

@@ -12,7 +12,7 @@ import json
 from collections import Counter
 from fastcore.utils import *
 from fastcore.meta import *
-from fastspec.errors import api_error_from_event
+from fasttransport.errors import api_error_from_event
 
 from .types import *
 from aidialog.msg_parts import (Part, PartType, Msg, Text, Thinking, ToolUse, ToolResult, tool_text,

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/openai_chat.py

@@ -12,7 +12,7 @@ import json
 from collections import Counter
 from fastcore.utils import *
 from fastcore.meta import *
-from fastspec.errors import api_error_from_event
+from fasttransport.errors import api_error_from_event
 
 from .types import *
 from aidialog.msg_parts import (Part, PartType, Msg, Text, Thinking, Refusal, ToolUse, ToolResult, tool_text,

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/openai_responses.py

@@ -12,7 +12,7 @@ import json
 from collections import Counter
 from fastcore.utils import *
 from fastcore.meta import *
-from fastspec.errors import api_error_from_event
+from fasttransport.errors import api_error_from_event
 
 from .types import *
 from aidialog.msg_parts import (Completion, Part, PartType, Msg, Text, Thinking, Refusal, ToolUse, ToolResult, tool_text,

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/responses.py

@@ -17,7 +17,7 @@ __all__ = ['ResponsesError', 'response_input', 'normalize_call_ids', 'response_o
 # %% ../nbs/06a_responses.ipynb #e16fc24e
 import asyncio, json, secrets, time
 from fastcore.utils import *
-from fastspec.errors import APIError
+from fasttransport.errors import APIError
 
 from .acomplete import acomplete
 from .streaming import Status

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/streaming.py

@@ -10,7 +10,7 @@ __all__ = ['Delta', 'norm_and_yield', 'PrintStream', 'PartAccum', 'FenceToolStop
 import json,copy
 from fastcore.utils import *
 from fastcore.meta import delegates
-from fastspec.errors import *
+from fasttransport.errors import *
 from .types import *
 from aidialog.msg_parts import Part, PartType, Msg, Text, Thinking, ToolUse, ToolResult, ServerToolResult, mk_part, fence_call_re, Completion
 

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/fastllm/types.py

@@ -4,17 +4,17 @@
 
 # %% auto #0
 __all__ = ['FinishReason', 'api_registry', 'model_prices_url', 'haik45', 'sonn45', 'sonn46', 'sonn', 'sonn5', 'opus46', 'opus48',
-           'opus', 'opus5', 'fable', 'fable5', 'gpt54', 'gpt54m', 'gpt55', 'codex54', 'codex54m', 'codex55',
-           'codex53spark', 'model_info_registry', 'modern_llm', 'deepseek_v4_common', 'mimo_v25_common',
-           'codex_pricing', 'sol', 'terra', 'luna', 'gpt56s', 'Usage', 'APIRegistry', 'mk_completion', 'fn_schema',
-           'payload_kwargs', 'provider_req', 'get_api_key', 'wrap_typed', 'unwrap_typed', 'resize_b64',
-           'model_prices_meta', 'infer_api_name', 'get_model_meta', 'register_model_info', 'get_model_info',
-           'get_model_pricing', 'approx_pricing', 'is_deepseek_peak_hour', 'price_tier', 'tier_rate']
+           'opus', 'opus5', 'fable', 'fable5', 'fable51', 'gpt54', 'gpt54m', 'gpt55', 'codex54', 'codex54m', 'codex55',
+           'codex53spark', 'model_info_registry', 'modern_llm', 'deepseek_v4_common', 'deepseek_v4_flash_prices',
+           'mimo_v25_common', 'codex_pricing', 'sol', 'terra', 'luna', 'gpt56s', 'Usage', 'APIRegistry',
+           'mk_completion', 'fn_schema', 'payload_kwargs', 'provider_req', 'get_api_key', 'wrap_typed', 'unwrap_typed',
+           'resize_b64', 'model_prices_meta', 'infer_api_name', 'get_model_meta', 'register_model_info',
+           'get_model_info', 'get_model_pricing', 'approx_pricing', 'is_deepseek_peak_hour', 'price_tier', 'tier_rate']
 
 # %% ../nbs/00_types.ipynb #b4d047fd
 import httpx2, base64, io
 from fasttransport.core import AsyncHttpCli
-import fastspec.errors  # patches httpx2 exceptions with .api_error(), which provider_req relies on
+import fasttransport.errors  # patches httpx2 exceptions with .api_error(), which provider_req relies on
 from importlib.metadata import entry_points
 from datetime import datetime, timezone
 from fastcore.utils import *
@@ -186,6 +186,7 @@ opus46 = "claude-opus-4-6"
 opus48 = "claude-opus-4-8"
 opus = opus5 = "claude-opus-5"
 fable = fable5 = 'claude-fable-5'
+fable51 = 'claude-fable-5-1'
 gpt54 = "gpt-5.4"
 gpt54m = "gpt-5.4-mini"
 gpt55 = "gpt-5.5"
@@ -258,14 +259,18 @@ register_model_info('kimi-k3', vendor_name='moonshot', base='kimi-k2.6', support
     max_input_tokens=1_000_000, input_cost_per_token=3e-6, cache_read_input_token_cost=0.30e-6, output_cost_per_token=15.0e-6)
 
 # %% ../nbs/00_types.ipynb #948d55d0
-deepseek_v4_common = dict(**modern_llm, max_input_tokens=1048576, max_output_tokens=393216, max_tokens=393216)
-
-register_model_info('deepseek-v4-flash', vendor_name='deepseek', base='deepseek/deepseek-v3.2', **deepseek_v4_common,
-    input_cost_per_token=1.4e-07, input_cost_per_token_cache_hit=2.8e-09,
-    output_cost_per_token=2.8e-07, cache_read_input_token_cost=1.4e-07/10)
-register_model_info('deepseek-v4-pro', vendor_name='deepseek', base='deepseek/deepseek-v3.2', **deepseek_v4_common,
-    input_cost_per_token=4.35e-07, input_cost_per_token_cache_hit=3.625e-09,
-    output_cost_per_token=8.7e-07, cache_read_input_token_cost=4.35e-07/10)
+# The price map carries DeepSeek's peak rates, and `Completion.cost` doubles at peak, so register the off-peak rates
+deepseek_v4_common = dict(supports_native_structured_output=True, max_input_tokens=1048576)
+deepseek_v4_flash_prices = dict(input_cost_per_token=0.22e-6, input_cost_per_token_cache_hit=0.007e-6,
+    cache_read_input_token_cost=0.007e-6, output_cost_per_token=0.66e-6)
+
+register_model_info('deepseek-v4-flash', vendor_name='deepseek', base='deepseek-v4-flash',
+    **deepseek_v4_common, **deepseek_v4_flash_prices)
+register_model_info('deepseek-v4-flash-vision-exp', vendor_name='deepseek', base='deepseek-v4-flash-vision-exp',
+    **deepseek_v4_common, **deepseek_v4_flash_prices, supports_image_input=True)
+register_model_info('deepseek-v4-pro', vendor_name='deepseek', base='deepseek-v4-pro', **deepseek_v4_common,
+    input_cost_per_token=0.66e-6, input_cost_per_token_cache_hit=0.022e-6,
+    cache_read_input_token_cost=0.022e-6, output_cost_per_token=1.98e-6)
 
 mimo_v25_common = dict(**modern_llm, supports_web_search=True, max_input_tokens=1048576, max_output_tokens=131072, max_tokens=131072)
 
@@ -310,7 +315,9 @@ register_model_info(codex53spark, 'codex', **codex_pricing,
     max_tokens=128000, max_input_tokens=128000, max_output_tokens=128000)
 
 # %% ../nbs/00_types.ipynb #5d3e4720
-for model in (haik45, sonn45, sonn46, sonn5, opus46, opus48, opus5, fable5):
+# Fable 5.1 isn't in the price map yet; it prices and behaves like Fable 5
+register_model_info(fable51, 'anthropic', base=fable5)
+for model in (haik45, sonn45, sonn46, sonn5, opus46, opus48, opus5, fable5, fable51):
     register_model_info(model, 'claude_code', base=model, base_vendor_name='anthropic', **codex_pricing)
 
 # %% ../nbs/00_types.ipynb #bb0c4c2a

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/pyproject.toml

@@ -15,7 +15,7 @@ classifiers = [
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3 :: Only",
 ]
-dependencies = ['fastcore>=2.1.18', 'aidialog>=0.0.10', 'fastspec>=0.2.1', 'fasttransport>=0.0.2', 'pillow']
+dependencies = ['fastcore>=2.1.18', 'aidialog>=0.0.10', 'fasttransport>=0.0.2', 'pillow']
 
 [project.urls]
 Repository = "https://github.com/AnswerDotAI/fastllm"

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/python_fastllm.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-fastllm
-Version: 0.0.44
+Version: 0.0.46
 Author-email: Kerem Turgutlu <keremturgutlu@gmail.com>
 License: Apache-2.0
 Project-URL: Repository, https://github.com/AnswerDotAI/fastllm
@@ -11,7 +11,6 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 Requires-Dist: fastcore>=2.1.18
 Requires-Dist: aidialog>=0.0.10
-Requires-Dist: fastspec>=0.2.1
 Requires-Dist: fasttransport>=0.0.2
 Requires-Dist: pillow
 

{python_fastllm-0.0.44 → python_fastllm-0.0.46}/python_fastllm.egg-info/requires.txt

@@ -1,5 +1,4 @@
 fastcore>=2.1.18
 aidialog>=0.0.10
-fastspec>=0.2.1
 fasttransport>=0.0.2
 pillow