scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-08-28 05:25:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 36

feat(gemini): day-0 pricing for gemini-3.6-flash and gemini-3.5-flash-lite

Browse source
This commit is contained in:
mateo-berri 2026-07-20 15:32:36 -07:00
parent c1b6c4062e
commit e411d637b3
3 changed files with 738 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

333
litellm/model_prices_and_context_window_backup.json
View file

@ -17566,6 +17566,61 @@
},
"web_search_billing_unit": "per_query"
},
"gemini-3.5-flash-lite": {
"cache_read_input_token_cost": 3e-08,
"cache_read_input_token_cost_flex": 2e-08,
"cache_read_input_token_cost_priority": 5e-08,
"input_cost_per_token": 3e-07,
"input_cost_per_token_batches": 1.5e-07,
"input_cost_per_token_flex": 1.5e-07,
"input_cost_per_token_priority": 5.4e-07,
"litellm_provider": "vertex_ai-language-models",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 2.5e-06,
"output_cost_per_token": 2.5e-06,
"output_cost_per_token_batches": 1.25e-06,
"output_cost_per_token_flex": 1.25e-06,
"output_cost_per_token_priority": 4.5e-06,
"source": "https://ai.google.dev/gemini-api/docs/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_audio_output": false,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"deep-research-pro-preview-12-2025": {
"input_cost_per_image": 0.0011,
"input_cost_per_token": 2e-06,
@ -18233,6 +18288,60 @@
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/gemini-3.6-flash": {
"cache_read_input_token_cost": 1.5e-07,
"cache_read_input_token_cost_flex": 7.5e-08,
"input_cost_per_token": 1.5e-06,
"input_cost_per_token_batches": 7.5e-07,
"input_cost_per_token_flex": 7.5e-07,
"litellm_provider": "vertex_ai",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 7.5e-06,
"output_cost_per_token": 7.5e-06,
"output_cost_per_token_batches": 3.75e-06,
"output_cost_per_token_flex": 3.75e-06,
"source": "https://cloud.google.com/vertex-ai/generative-ai/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"input_cost_per_token_priority": 2.7e-06,
"output_cost_per_token_priority": 1.35e-05,
"cache_read_input_token_cost_priority": 2.7e-07,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/gemini-3.1-pro-preview": {
"cache_read_input_token_cost": 2e-07,
"cache_read_input_token_cost_above_200k_tokens": 4e-07,
@ -19585,6 +19694,63 @@
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-3.5-flash-lite": {
"cache_read_input_token_cost": 3e-08,
"cache_read_input_token_cost_flex": 2e-08,
"cache_read_input_token_cost_priority": 5e-08,
"input_cost_per_token": 3e-07,
"input_cost_per_token_batches": 1.5e-07,
"input_cost_per_token_flex": 1.5e-07,
"input_cost_per_token_priority": 5.4e-07,
"litellm_provider": "gemini",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 2.5e-06,
"output_cost_per_token": 2.5e-06,
"output_cost_per_token_batches": 1.25e-06,
"output_cost_per_token_flex": 1.25e-06,
"output_cost_per_token_priority": 4.5e-06,
"rpm": 15,
"source": "https://ai.google.dev/gemini-api/docs/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_audio_output": false,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"tpm": 250000,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-3-flash-preview": {
"cache_read_input_token_cost": 5e-08,
"input_cost_per_audio_token": 1e-06,
@ -19691,6 +19857,63 @@
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-3.6-flash": {
"cache_read_input_token_cost": 1.5e-07,
"cache_read_input_token_cost_flex": 7.5e-08,
"input_cost_per_token": 1.5e-06,
"input_cost_per_token_batches": 7.5e-07,
"input_cost_per_token_flex": 7.5e-07,
"litellm_provider": "gemini",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 7.5e-06,
"output_cost_per_token": 7.5e-06,
"output_cost_per_token_batches": 3.75e-06,
"output_cost_per_token_flex": 3.75e-06,
"rpm": 2000,
"source": "https://ai.google.dev/pricing/gemini-3",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_output": false,
"supports_audio_input": true,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"tpm": 800000,
"input_cost_per_token_priority": 2.7e-06,
"output_cost_per_token_priority": 1.35e-05,
"cache_read_input_token_cost_priority": 2.7e-07,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-omni-flash-preview": {
"input_cost_per_audio_token": 1.5e-06,
"input_cost_per_token": 1.5e-06,
@ -19971,6 +20194,61 @@
},
"web_search_billing_unit": "per_query"
},
"gemini-3.6-flash": {
"cache_read_input_token_cost": 1.5e-07,
"cache_read_input_token_cost_flex": 7.5e-08,
"input_cost_per_token": 1.5e-06,
"input_cost_per_token_batches": 7.5e-07,
"input_cost_per_token_flex": 7.5e-07,
"litellm_provider": "vertex_ai-language-models",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 7.5e-06,
"output_cost_per_token": 7.5e-06,
"output_cost_per_token_batches": 3.75e-06,
"output_cost_per_token_flex": 3.75e-06,
"source": "https://ai.google.dev/pricing/gemini-3",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_output": false,
"supports_audio_input": true,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"input_cost_per_token_priority": 2.7e-06,
"output_cost_per_token_priority": 1.35e-05,
"cache_read_input_token_cost_priority": 2.7e-07,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-2.5-pro-preview-tts": {
"cache_read_input_token_cost": 1.25e-07,
"cache_read_input_token_cost_above_200k_tokens": 2.5e-07,
@ -37232,6 +37510,61 @@
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/gemini-3.5-flash-lite": {
"cache_read_input_token_cost": 3e-08,
"cache_read_input_token_cost_flex": 2e-08,
"cache_read_input_token_cost_priority": 5e-08,
"input_cost_per_token": 3e-07,
"input_cost_per_token_batches": 1.5e-07,
"input_cost_per_token_flex": 1.5e-07,
"input_cost_per_token_priority": 5.4e-07,
"litellm_provider": "vertex_ai-language-models",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 2.5e-06,
"output_cost_per_token": 2.5e-06,
"output_cost_per_token_batches": 1.25e-06,
"output_cost_per_token_flex": 1.25e-06,
"output_cost_per_token_priority": 4.5e-06,
"source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_audio_output": false,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/deep-research-pro-preview-12-2025": {
"input_cost_per_image": 0.0011,
"input_cost_per_token": 2e-06,

333
model_prices_and_context_window.json
View file

@ -17644,6 +17644,61 @@
},
"web_search_billing_unit": "per_query"
},
"gemini-3.5-flash-lite": {
"cache_read_input_token_cost": 3e-08,
"cache_read_input_token_cost_flex": 2e-08,
"cache_read_input_token_cost_priority": 5e-08,
"input_cost_per_token": 3e-07,
"input_cost_per_token_batches": 1.5e-07,
"input_cost_per_token_flex": 1.5e-07,
"input_cost_per_token_priority": 5.4e-07,
"litellm_provider": "vertex_ai-language-models",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 2.5e-06,
"output_cost_per_token": 2.5e-06,
"output_cost_per_token_batches": 1.25e-06,
"output_cost_per_token_flex": 1.25e-06,
"output_cost_per_token_priority": 4.5e-06,
"source": "https://ai.google.dev/gemini-api/docs/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_audio_output": false,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"deep-research-pro-preview-12-2025": {
"input_cost_per_image": 0.0011,
"input_cost_per_token": 2e-06,
@ -18311,6 +18366,60 @@
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/gemini-3.6-flash": {
"cache_read_input_token_cost": 1.5e-07,
"cache_read_input_token_cost_flex": 7.5e-08,
"input_cost_per_token": 1.5e-06,
"input_cost_per_token_batches": 7.5e-07,
"input_cost_per_token_flex": 7.5e-07,
"litellm_provider": "vertex_ai",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 7.5e-06,
"output_cost_per_token": 7.5e-06,
"output_cost_per_token_batches": 3.75e-06,
"output_cost_per_token_flex": 3.75e-06,
"source": "https://cloud.google.com/vertex-ai/generative-ai/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"input_cost_per_token_priority": 2.7e-06,
"output_cost_per_token_priority": 1.35e-05,
"cache_read_input_token_cost_priority": 2.7e-07,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/gemini-3.1-pro-preview": {
"cache_read_input_token_cost": 2e-07,
"cache_read_input_token_cost_above_200k_tokens": 4e-07,
@ -19663,6 +19772,63 @@
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-3.5-flash-lite": {
"cache_read_input_token_cost": 3e-08,
"cache_read_input_token_cost_flex": 2e-08,
"cache_read_input_token_cost_priority": 5e-08,
"input_cost_per_token": 3e-07,
"input_cost_per_token_batches": 1.5e-07,
"input_cost_per_token_flex": 1.5e-07,
"input_cost_per_token_priority": 5.4e-07,
"litellm_provider": "gemini",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 2.5e-06,
"output_cost_per_token": 2.5e-06,
"output_cost_per_token_batches": 1.25e-06,
"output_cost_per_token_flex": 1.25e-06,
"output_cost_per_token_priority": 4.5e-06,
"rpm": 15,
"source": "https://ai.google.dev/gemini-api/docs/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_audio_output": false,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"tpm": 250000,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-3-flash-preview": {
"cache_read_input_token_cost": 5e-08,
"input_cost_per_audio_token": 1e-06,
@ -19769,6 +19935,63 @@
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-3.6-flash": {
"cache_read_input_token_cost": 1.5e-07,
"cache_read_input_token_cost_flex": 7.5e-08,
"input_cost_per_token": 1.5e-06,
"input_cost_per_token_batches": 7.5e-07,
"input_cost_per_token_flex": 7.5e-07,
"litellm_provider": "gemini",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 7.5e-06,
"output_cost_per_token": 7.5e-06,
"output_cost_per_token_batches": 3.75e-06,
"output_cost_per_token_flex": 3.75e-06,
"rpm": 2000,
"source": "https://ai.google.dev/pricing/gemini-3",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_output": false,
"supports_audio_input": true,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"tpm": 800000,
"input_cost_per_token_priority": 2.7e-06,
"output_cost_per_token_priority": 1.35e-05,
"cache_read_input_token_cost_priority": 2.7e-07,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-omni-flash-preview": {
"input_cost_per_audio_token": 1.5e-06,
"input_cost_per_token": 1.5e-06,
@ -20049,6 +20272,61 @@
},
"web_search_billing_unit": "per_query"
},
"gemini-3.6-flash": {
"cache_read_input_token_cost": 1.5e-07,
"cache_read_input_token_cost_flex": 7.5e-08,
"input_cost_per_token": 1.5e-06,
"input_cost_per_token_batches": 7.5e-07,
"input_cost_per_token_flex": 7.5e-07,
"litellm_provider": "vertex_ai-language-models",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 7.5e-06,
"output_cost_per_token": 7.5e-06,
"output_cost_per_token_batches": 3.75e-06,
"output_cost_per_token_flex": 3.75e-06,
"source": "https://ai.google.dev/pricing/gemini-3",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_output": false,
"supports_audio_input": true,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"input_cost_per_token_priority": 2.7e-06,
"output_cost_per_token_priority": 1.35e-05,
"cache_read_input_token_cost_priority": 2.7e-07,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"gemini/gemini-2.5-pro-preview-tts": {
"cache_read_input_token_cost": 1.25e-07,
"cache_read_input_token_cost_above_200k_tokens": 2.5e-07,
@ -37323,6 +37601,61 @@
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/gemini-3.5-flash-lite": {
"cache_read_input_token_cost": 3e-08,
"cache_read_input_token_cost_flex": 2e-08,
"cache_read_input_token_cost_priority": 5e-08,
"input_cost_per_token": 3e-07,
"input_cost_per_token_batches": 1.5e-07,
"input_cost_per_token_flex": 1.5e-07,
"input_cost_per_token_priority": 5.4e-07,
"litellm_provider": "vertex_ai-language-models",
"max_input_tokens": 1048576,
"max_output_tokens": 65536,
"max_tokens": 65536,
"mode": "chat",
"output_cost_per_reasoning_token": 2.5e-06,
"output_cost_per_token": 2.5e-06,
"output_cost_per_token_batches": 1.25e-06,
"output_cost_per_token_flex": 1.25e-06,
"output_cost_per_token_priority": 4.5e-06,
"source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/completions",
"/v1/batch"
],
"supported_modalities": [
"text",
"image",
"audio",
"video"
],
"supported_output_modalities": [
"text"
],
"supports_audio_input": true,
"supports_audio_output": false,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true,
"supports_url_context": true,
"supports_video_input": true,
"supports_vision": true,
"supports_web_search": true,
"supports_native_streaming": true,
"search_context_cost_per_query": {
"search_context_size_low": 0.014,
"search_context_size_medium": 0.014,
"search_context_size_high": 0.014
},
"web_search_billing_unit": "per_query"
},
"vertex_ai/deep-research-pro-preview-12-2025": {
"input_cost_per_image": 0.0011,
"input_cost_per_token": 2e-06,

72
tests/test_litellm/litellm_core_utils/llm_cost_calc/test_llm_cost_calc_utils.py
View file

@ -2237,3 +2237,75 @@ def test_token_type_cost_breakdown_applies_regional_uplift():
text_input_cost = 600 * model_info["input_cost_per_token"] * uplift
assert text_output_cost + eu.reasoning_cost == pytest.approx(completion_cost)
assert text_input_cost + eu.cache_read_cost == pytest.approx(prompt_cost)
GEMINI_DAY0_LAUNCH_PRICING = [
("gemini-3.6-flash", 1.5e-06, 7.5e-06, 1.5e-07),
("gemini/gemini-3.6-flash", 1.5e-06, 7.5e-06, 1.5e-07),
("vertex_ai/gemini-3.6-flash", 1.5e-06, 7.5e-06, 1.5e-07),
("gemini-3.5-flash-lite", 3e-07, 2.5e-06, 3e-08),
("gemini/gemini-3.5-flash-lite", 3e-07, 2.5e-06, 3e-08),
("vertex_ai/gemini-3.5-flash-lite", 3e-07, 2.5e-06, 3e-08),
]
@pytest.mark.parametrize("model,input_cost,output_cost,cache_read_cost", GEMINI_DAY0_LAUNCH_PRICING)
def test_gemini_36_flash_and_35_flash_lite_launch_pricing(model, input_cost, output_cost, cache_read_cost):
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
model_cost_map = litellm.model_cost[model]
assert model_cost_map["input_cost_per_token"] == input_cost
assert model_cost_map["output_cost_per_token"] == output_cost
assert model_cost_map["output_cost_per_reasoning_token"] == output_cost
assert model_cost_map["cache_read_input_token_cost"] == cache_read_cost
assert model_cost_map["mode"] == "chat"
assert model_cost_map["supports_reasoning"] is True
assert model_cost_map["supports_function_calling"] is True
assert model_cost_map["max_input_tokens"] == 1048576
def test_generic_cost_per_token_gemini_36_flash():
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
usage = Usage(
prompt_tokens=1000,
completion_tokens=500,
total_tokens=1500,
completion_tokens_details=CompletionTokensDetailsWrapper(
reasoning_tokens=200,
text_tokens=300,
),
prompt_tokens_details=PromptTokensDetailsWrapper(text_tokens=1000),
)
prompt_cost, completion_cost = generic_cost_per_token(
model="gemini-3.6-flash",
usage=usage,
custom_llm_provider="gemini",
)
assert prompt_cost == pytest.approx(0.0015)
assert completion_cost == pytest.approx(0.00375)
def test_generic_cost_per_token_gemini_35_flash_lite():
os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True"
litellm.model_cost = litellm.get_model_cost_map(url="")
usage = Usage(
prompt_tokens=1000,
completion_tokens=500,
total_tokens=1500,
completion_tokens_details=CompletionTokensDetailsWrapper(
reasoning_tokens=200,
text_tokens=300,
),
prompt_tokens_details=PromptTokensDetailsWrapper(text_tokens=1000),
)
prompt_cost, completion_cost = generic_cost_per_token(
model="gemini-3.5-flash-lite",
usage=usage,
custom_llm_provider="gemini",
)
assert prompt_cost == pytest.approx(0.0003)
assert completion_cost == pytest.approx(0.00125)