scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-11-07 07:58:09 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 150

fix(pricing): price Gemini Live image input at Vertex's non-text rate

Browse source 
The new Live entries carried no input_cost_per_image_token, so image tokens fell
through to input_cost_per_token and billed at the $0.50/M text rate instead of
$3.00/M, a 6x under-bill.

Image input matters more on Live than the name suggests: Gemini reports a Live
video stream as a sequence of frames under the IMAGE modality, so a camera feed
is billed entirely through this field, and VIDEO never appears in a Live
session's usage. Google documents video input as "roughly 258 tokens per second",
and a single JPEG frame measured exactly 258 tokens against Vertex, so one frame
is one second of video for billing.

$3.00 per 1M tokens is Google's non-text Live input rate, on
ai.google.dev/gemini-api/docs/pricing as "$3.00 (audio / video)" and on the Vertex
generative-ai price sheet, matching the audio rate already in these entries.

Applied to both the bare and the vertex_ai/-prefixed keys, since the prefixed one
is what get_model_info resolves for custom_llm_provider="vertex_ai".
This commit is contained in:
Marty Sullivan 2026-08-14 20:05:16 -04:00
parent 19249242f8
commit 9c538a0b2f
2 changed files with 8 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
litellm/model_prices_and_context_window_backup.json
View file

@ -18472,6 +18472,7 @@
"gemini-live-2.5-flash": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,
@ -18519,6 +18520,7 @@
"vertex_ai/gemini-live-2.5-flash": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,
@ -18566,6 +18568,7 @@
"gemini-live-2.5-flash-native-audio": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,
@ -18623,6 +18626,7 @@
"vertex_ai/gemini-live-2.5-flash-native-audio": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,

4
model_prices_and_context_window.json
View file

@ -18472,6 +18472,7 @@
"gemini-live-2.5-flash": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,
@ -18519,6 +18520,7 @@
"vertex_ai/gemini-live-2.5-flash": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,
@ -18566,6 +18568,7 @@
"gemini-live-2.5-flash-native-audio": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,
@ -18623,6 +18626,7 @@
"vertex_ai/gemini-live-2.5-flash-native-audio": {
"cache_read_input_token_cost": 7.5e-08,
"input_cost_per_audio_token": 3e-06,
"input_cost_per_image_token": 3e-06,
"input_cost_per_token": 5e-07,
"output_cost_per_audio_token": 1.2e-05,
"output_cost_per_token": 2e-06,