chore(pricing): Update google pricing

[siddharthsambharia-portkey]

🔄 Pricing Update: google

📊 Summary (complete_diff mode)

Change Type	Count
➕ Models added	4
🔄 Models updated (merged)	36

➕ New Models

• gemini-embedding-2-preview-lte-128k
• gemini-embedding-2-preview-gt-128k
• veo-3.1-lite-generate-preview-lte-128k
• veo-3.1-lite-generate-preview-gt-128k

🔄 Updated Models

• gemini-2.5-pro-lte-128k
• gemini-2.5-pro-gt-128k
• gemini-2.5-flash-lte-128k
• gemini-2.5-flash-gt-128k
• gemini-2.5-flash-lite-lte-128k
• gemini-2.5-flash-lite-gt-128k
• gemini-2.0-flash-lte-128k
• gemini-2.0-flash-gt-128k
• gemini-2.0-flash-001-lte-128k
• gemini-2.0-flash-001-gt-128k
• gemini-2.0-flash-lite-lte-128k
• gemini-2.0-flash-lite-gt-128k
• gemini-2.0-flash-lite-001-lte-128k
• gemini-2.0-flash-lite-001-gt-128k
• gemini-3-pro-preview-lte-128k
• gemini-3-pro-preview-gt-128k
• gemini-3-flash-preview-lte-128k
• gemini-3-flash-preview-gt-128k
• gemini-3.1-pro-preview-lte-128k
• gemini-3.1-pro-preview-gt-128k
• gemini-3.1-pro-preview-customtools-lte-128k
• gemini-3.1-pro-preview-customtools-gt-128k
• gemini-3.1-flash-lite-preview-lte-128k
• gemini-3.1-flash-lite-preview-gt-128k
• gemini-2.5-flash-image-lte-128k
• gemini-2.5-flash-image-gt-128k
• gemini-3.1-flash-image-preview-lte-128k
• gemini-3.1-flash-image-preview-gt-128k
• gemini-3-pro-image-preview-lte-128k
• gemini-3-pro-image-preview-gt-128k
• ... and 6 more

Data Sources

• Google Gemini Models API — fetched via get_gemini_models (50 models total)
• LiteLLM Model Prices JSON (https://raw.githubusercontent.com/BerriAI/litellm/main/model_prices_and_context_window.json) — used as pricing proxy because Firecrawl credits were exhausted and the Google pricing page (~194k chars) exceeded all tool read limits

Note: firecrawl_scrape returned "Insufficient credits". Multiple attempts to read the raw HTTP response file also failed (194k-char single line exceeds tool limits). The LiteLLM JSON was the only accessible source that returned pricing data directly.

---

Model → Pricing Page Mapping

Model ID	Pricing source / section	Notes
gemini-2.5-pro-lte-128k	LiteLLM: gemini/gemini-2.5-pro (≤200k tier)	input $1.25/1M, output $10/1M, cache_read $0.125/1M, thinking_token $3.5/1M, batch 50%, web_search
gemini-2.5-pro-gt-128k	LiteLLM: gemini/gemini-2.5-pro (>200k tier)	input $2.50/1M, output $15/1M, cache_read $0.25/1M, thinking_token $3.5/1M, batch 50%, web_search
gemini-2.5-flash-lte-128k	LiteLLM: gemini/gemini-2.5-flash	input $0.30/1M, output $2.50/1M, cache_read $0.03/1M, batch 50%, web_search; flat pricing (no 128k tier on page)
gemini-2.5-flash-gt-128k	LiteLLM: gemini/gemini-2.5-flash	same as lte (flat pricing)
gemini-2.5-flash-lite-lte-128k	LiteLLM: gemini/gemini-2.5-flash-lite	input $0.10/1M, output $0.40/1M, cache_read $0.01/1M, web_search
gemini-2.5-flash-lite-gt-128k	LiteLLM: gemini/gemini-2.5-flash-lite	same as lte (flat pricing)
gemini-2.0-flash-lte-128k	LiteLLM: gemini/gemini-2.0-flash	input $0.10/1M, output $0.40/1M, cache_read $0.025/1M, web_search
gemini-2.0-flash-gt-128k	LiteLLM: gemini/gemini-2.0-flash	same as lte (flat pricing)
gemini-2.0-flash-001-lte-128k	LiteLLM: gemini/gemini-2.0-flash-001	same pricing as gemini-2.0-flash
gemini-2.0-flash-001-gt-128k	LiteLLM: gemini/gemini-2.0-flash-001	same as lte
gemini-2.0-flash-lite-lte-128k	LiteLLM: gemini/gemini-2.0-flash-lite	input $0.075/1M, output $0.30/1M, cache_read $0.01875/1M, web_search
gemini-2.0-flash-lite-gt-128k	LiteLLM: gemini/gemini-2.0-flash-lite	same as lte
gemini-2.0-flash-lite-001-lte-128k	LiteLLM: gemini/gemini-2.0-flash-lite-001	same pricing as gemini-2.0-flash-lite
gemini-2.0-flash-lite-001-gt-128k	LiteLLM: gemini/gemini-2.0-flash-lite-001	same as lte
gemini-3-pro-preview-lte-128k	LiteLLM: gemini/gemini-3-pro-preview (≤200k)	input $2.00/1M, output $12/1M, cache_read $0.20/1M, batch $1/$6, web_search
gemini-3-pro-preview-gt-128k	LiteLLM: gemini/gemini-3-pro-preview (>200k)	input $4.00/1M, output $18/1M, cache_read $0.40/1M, batch $2/$9, web_search
gemini-3-flash-preview-lte-128k	LiteLLM: gemini/gemini-3-flash-preview	input $0.50/1M, output $3.00/1M, cache_read $0.05/1M, web_search; thinking in output price
gemini-3-flash-preview-gt-128k	LiteLLM: gemini/gemini-3-flash-preview	same as lte (flat pricing)
gemini-3.1-pro-preview-lte-128k	LiteLLM: gemini/gemini-3.1-pro-preview (≤200k)	input $2.00/1M, output $12/1M, cache_read $0.20/1M, batch $1/$6, web_search
gemini-3.1-pro-preview-gt-128k	LiteLLM: gemini/gemini-3.1-pro-preview (>200k)	input $4.00/1M, output $18/1M, cache_read $0.40/1M, batch $2/$9, web_search
gemini-3.1-pro-preview-customtools-lte-128k	LiteLLM: gemini/gemini-3.1-pro-preview-customtools	same pricing as gemini-3.1-pro-preview
gemini-3.1-pro-preview-customtools-gt-128k	LiteLLM: gemini/gemini-3.1-pro-preview-customtools	same as lte variant for customtools
gemini-3.1-flash-lite-preview-lte-128k	LiteLLM: gemini/gemini-3.1-flash-lite-preview	input $0.25/1M, output $1.50/1M, cache_read $0.025/1M, web_search
gemini-3.1-flash-lite-preview-gt-128k	LiteLLM: gemini/gemini-3.1-flash-lite-preview	same as lte (flat pricing)
gemini-2.5-flash-image-lte-128k	LiteLLM: gemini/gemini-2.5-flash-image	input $0.30/1M, text output $2.50/1M, image_token $30/1M, cache_read $0.03/1M, batch $0.15/$1.25, web_search
gemini-2.5-flash-image-gt-128k	LiteLLM: gemini/gemini-2.5-flash-image	same as lte (flat pricing for image models)
gemini-3.1-flash-image-preview-lte-128k	LiteLLM: gemini/gemini-3.1-flash-image-preview	input $0.25/1M, text output $1.50/1M, image_token $60/1M, batch $0.125/$0.75, web_search; batch image rate $30/1M
gemini-3.1-flash-image-preview-gt-128k	LiteLLM: gemini/gemini-3.1-flash-image-preview	same as lte (flat pricing)
gemini-3-pro-image-preview-lte-128k	LiteLLM: gemini/gemini-3-pro-image-preview	input $2.00/1M, text output $12/1M, image_token $120/1M, batch $1/$6, web_search; flat pricing
gemini-3-pro-image-preview-gt-128k	LiteLLM: gemini/gemini-3-pro-image-preview	same as lte (explicitly flat pricing)
gemini-flash-latest-lte-128k	LiteLLM: gemini/gemini-flash-latest → resolves to gemini-2.5-flash pricing	input $0.30/1M, output $2.50/1M, cache_read $0.03/1M; NOTE: skill table suggests this resolves to gemini-3-flash-preview, but LiteLLM maps it to 2.5-flash series — could not verify from live pricing page
gemini-flash-latest-gt-128k	LiteLLM: gemini/gemini-flash-latest	same as lte
gemini-flash-lite-latest-lte-128k	LiteLLM: gemini/gemini-flash-lite-latest → resolves to gemini-2.5-flash-lite pricing	input $0.10/1M, output $0.40/1M, cache_read $0.01/1M; NOTE: skill table suggests gemini-3.1-flash-lite-preview, but LiteLLM maps to 2.5 — could not verify from live page
gemini-flash-lite-latest-gt-128k	LiteLLM: gemini/gemini-flash-lite-latest	same as lte
gemini-pro-latest-lte-128k	LiteLLM: gemini/gemini-pro-latest → resolves to gemini-2.5-pro pricing (≤200k tier)	input $1.25/1M, output $10/1M, cache_read $0.125/1M; NOTE: skill table suggests gemini-3.1-pro-preview, but LiteLLM maps to 2.5-pro — could not verify from live page
gemini-pro-latest-gt-128k	LiteLLM: gemini/gemini-pro-latest (>200k tier)	input $2.50/1M, output $15/1M, cache_read $0.25/1M
gemini-embedding-001-lte-128k	LiteLLM: gemini/gemini-embedding-001	input $0.15/1M, output 0
gemini-embedding-001-gt-128k	LiteLLM: gemini/gemini-embedding-001	same as lte
gemini-embedding-2-preview-lte-128k	LiteLLM: gemini/gemini-embedding-2-preview	input $0.20/1M, output 0; multimodal embedding
gemini-embedding-2-preview-gt-128k	LiteLLM: gemini/gemini-embedding-2-preview	same as lte
imagen-4.0-generate-001-lte-128k	LiteLLM: gemini/imagen-4.0-generate-001	$0.04/image
imagen-4.0-generate-001-gt-128k	LiteLLM: gemini/imagen-4.0-generate-001	same as lte
imagen-4.0-ultra-generate-001-lte-128k	LiteLLM: gemini/imagen-4.0-ultra-generate-001	$0.06/image
imagen-4.0-ultra-generate-001-gt-128k	LiteLLM: gemini/imagen-4.0-ultra-generate-001	same as lte
imagen-4.0-fast-generate-001-lte-128k	LiteLLM: gemini/imagen-4.0-fast-generate-001	$0.02/image
imagen-4.0-fast-generate-001-gt-128k	LiteLLM: gemini/imagen-4.0-fast-generate-001	same as lte
veo-2.0-generate-001-lte-128k	LiteLLM: gemini/veo-2.0-generate-001	$0.35/s video, 8s default, 1 sample
veo-2.0-generate-001-gt-128k	LiteLLM: gemini/veo-2.0-generate-001	same as lte
veo-3.0-generate-001-lte-128k	LiteLLM: gemini/veo-3.0-generate-001	$0.40/s video, 8s default, 1 sample
veo-3.0-generate-001-gt-128k	LiteLLM: gemini/veo-3.0-generate-001	same as lte
veo-3.0-fast-generate-001-lte-128k	LiteLLM: gemini/veo-3.0-fast-generate-001	$0.15/s video, 8s default, 1 sample
veo-3.0-fast-generate-001-gt-128k	LiteLLM: gemini/veo-3.0-fast-generate-001	same as lte
veo-3.1-generate-preview-lte-128k	LiteLLM: gemini/veo-3.1-generate-preview	$0.40/s video, 8s default, 1 sample
veo-3.1-generate-preview-gt-128k	LiteLLM: gemini/veo-3.1-generate-preview	same as lte
veo-3.1-fast-generate-preview-lte-128k	LiteLLM: gemini/veo-3.1-fast-generate-preview	$0.15/s video, 8s default, 1 sample
veo-3.1-fast-generate-preview-gt-128k	LiteLLM: gemini/veo-3.1-fast-generate-preview	same as lte
veo-3.1-lite-generate-preview-lte-128k	Not in LiteLLM; price unknown	video_seconds=0 (price not found); needs manual verification
veo-3.1-lite-generate-preview-gt-128k	Not in LiteLLM; price unknown	same as lte with 0s

---

Excluded Models (per skill rules)

Model	Reason
gemini-2.5-flash-preview-tts, gemini-2.5-pro-preview-tts	TTS models — excluded
gemma-3-, gemma-4-, gemma-3n-*	Gemma family — excluded
nano-banana-pro-preview	Contains "nano" in model ID — excluded per nano rule
gemini-robotics-er-1.5-preview	Robotics — not in include list
lyria-3-clip-preview, lyria-3-pro-preview	Not in include list
gemini-2.5-computer-use-preview-10-2025	computer-use variant — excluded
deep-research-pro-preview-12-2025	Not in include list
gemini-2.5-flash-native-audio-latest/preview	native-audio — excluded
gemini-3.1-flash-live-preview	live variant — excluded
aqa	Not in include list

---

*-latest Alias Resolution Note

The skill reference table indicates:

• gemini-pro-latest → gemini-3.1-pro-preview
• gemini-flash-latest → gemini-3-flash-preview
• gemini-flash-lite-latest → gemini-3.1-flash-lite-preview

However, the LiteLLM JSON (the only accessible pricing source) maps these aliases to the 2.5 series. Without access to the live Google pricing page (Firecrawl credits exhausted), the 2.5-series pricing from LiteLLM was used. Reviewers should verify the correct alias resolution from the live pricing page.

---

Generated by Pricing Agent on 2026-04-05