python-fastllm 0.0.42__tar.gz → 0.0.43__tar.gz

{python_fastllm-0.0.42 → python_fastllm-0.0.43}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-fastllm
-Version: 0.0.42
+Version: 0.0.43
 Author-email: Kerem Turgutlu <keremturgutlu@gmail.com>
 License: Apache-2.0
 Project-URL: Repository, https://github.com/AnswerDotAI/fastllm

python_fastllm-0.0.43/fastllm/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.0.43"

{python_fastllm-0.0.42 → python_fastllm-0.0.43}/fastllm/acomplete.py

@@ -47,7 +47,8 @@ vendor_mapping = {
     "fireworks_ai": ('openai_chat', "https://api.fireworks.ai/inference/v1", "FIREWORKS_API_KEY"),
     "qwen":         ('openai_chat', "https://dashscope.aliyuncs.com/compatible-mode/v1", "QWEN_API_KEY"),
     "minimax":      ('anthropic', "https://api.minimax.io/anthropic", "MINIMAX_API_KEY"),
-    "meta_ai":      ('openai', "https://api.meta.ai/v1", "META_API_KEY")
+    "meta_ai":      ('openai', "https://api.meta.ai/v1", "META_API_KEY"),
+    "zai":          ('openai_chat', "https://api.z.ai/api/paas/v4", "ZAI_API_KEY")
 }
 
 # %% ../nbs/06_acomplete.ipynb #77d27ea7

{python_fastllm-0.0.42 → python_fastllm-0.0.43}/fastllm/types.py

@@ -265,6 +265,13 @@ register_model_info('muse-spark-1.1', vendor_name='meta_ai', **modern_llm, max_i
     supports_vision=True, supports_image_input=True, supports_video_input=True, supports_pdf_input=True, supports_web_search=True,
     search_context_cost_per_query=0.0025, input_cost_per_token=1.25e-6, output_cost_per_token=4.25e-6, cache_read_input_token_cost=0.15e-6)
 
+# %% ../nbs/00_types.ipynb #5a2b3430
+# z.ai list pricing (a 50% discount applies until 2026-09-09)
+register_model_info('glm-5.3-flash', vendor_name='zai', base='glm-5.1', **modern_llm,
+    supports_vision=True, supports_image_input=True,
+    max_input_tokens=1_000_000, max_output_tokens=128_000, max_tokens=128_000,
+    input_cost_per_token=0.15e-6, cache_read_input_token_cost=0.03e-6, output_cost_per_token=0.5e-6)
+
 # %% ../nbs/00_types.ipynb #2c23d11e
 codex_pricing = dict(input_cost_per_token = 0.10/1_000_000, output_cost_per_token = 0.50/1_000_000,
     cache_creation_input_token_cost = 0.10/1_000_000, cache_read_input_token_cost = 0.10/1_000_000)

{python_fastllm-0.0.42 → python_fastllm-0.0.43}/python_fastllm.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-fastllm
-Version: 0.0.42
+Version: 0.0.43
 Author-email: Kerem Turgutlu <keremturgutlu@gmail.com>
 License: Apache-2.0
 Project-URL: Repository, https://github.com/AnswerDotAI/fastllm

python_fastllm-0.0.42/fastllm/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.0.42"