chore(pricing): Update vertex-ai pricing

[siddharthsambharia-portkey]

🔄 Pricing Update: vertex-ai

📊 Summary (complete_diff mode)

Change Type	Count
➕ Models added	3
🔄 Models updated (merged)	13

➕ New Models

• gemini-2.5-pro-tts
• gemini-2.5-flash-tts
• veo-3.1-lite-generate-001

🔄 Updated Models

• gemini-2.5-pro
• gemini-2.5-computer-use-preview-10-2025
• gemini-2.0-flash-001
• gemini-3.1-pro-preview
• gemini-3.1-flash-lite-preview
• gemini-3.1-flash-image-preview
• gemini-3-pro-image-preview
• gemini-3-flash-preview
• veo-3.0-fast-generate-001
• veo-3.1-fast-generate-001
• gemini-embedding-001
• textembedding-gecko-multilingual@001
• multimodalembedding@001

Model-to-Pricing-Page Mapping

Model ID	Publisher / Section	Source	Notes
gemini-2.5-pro	Google – Gemini 2.5	API	Standard ≤200K pricing; long-context (>200K) $2.50/$15 noted
gemini-2.5-flash	Google – Gemini 2.5	API	Includes audio input $1.00/1M
gemini-2.5-flash-image	Google – Gemini 2.5	API	Gemini image-output variant; image_token $30/1M
gemini-2.5-flash-lite	Google – Gemini 2.5	API	Lightest 2.5 model
gemini-2.5-computer-use-preview-10-2025	Google – Gemini 2.5	API	Maps to Gemini 2.5 Pro pricing; no cache row found
gemini-2.5-flash-preview-09-2025	Google – Gemini 2.5	API	Preview alias; same pricing as gemini-2.5-flash
gemini-2.5-flash-lite-preview-09-2025	Google – Gemini 2.5	API	Preview alias; same pricing as gemini-2.5-flash-lite
gemini-2.0-flash-001	Google – Gemini 2.0	API	Includes audio input $1.00/1M; cache_read $0.0375/1M
gemini-2.0-flash-lite-001	Google – Gemini 2.0	API	No cache row on page
gemini-3.1-pro-preview	Google – Gemini 3.1	API	Standard ≤200K pricing; long-context (>200K) $4/$18 noted
gemini-3.1-flash-lite-preview	Google – Gemini 3.1	API	Includes audio input $0.50/1M
gemini-3.1-flash-image-preview	Google – Gemini 3.1	API	image_token $60/1M
gemini-3-pro-preview	Google – Gemini 3	API	Standard ≤200K pricing; long-context (>200K) $4/$18 noted
gemini-3-pro-image-preview	Google – Gemini 3	API	image_token $120/1M
gemini-3-flash-preview	Google – Gemini 3	API	Includes audio input $1.00/1M
gemini-2.5-pro-tts	Google – Gemini	API – price not found	No TTS pricing row on page; added with price 0
gemini-2.5-flash-tts	Google – Gemini	API – price not found	No TTS pricing row on page; added with price 0
imagen-4.0-generate-001	Google – Imagen 4	API	$0.04/image; row matched via lookup_variant imagen-4.0-generate
imagen-4.0-fast-generate-001	Google – Imagen 4	API	$0.02/image; row matched via lookup_variant imagen-4.0-fast-generate
imagen-4.0-ultra-generate-001	Google – Imagen 4	API	$0.06/image; row matched via lookup_variant imagen-4.0-ultra-generate
imagen-3.0-generate-002	Google – Imagen 3	API	$0.04/image; row matched via lookup_variant imagen-3.0-generate
imagen-3.0-capability-001	Google – Imagen 3	API	Capability model; uses imagen-3.0-generate price $0.04/image
imagen-3.0-capability-002	Google – Imagen 3	API	Capability model; uses imagen-3.0-generate price $0.04/image
veo-2.0-generate-001	Google – Veo 2	API	$0.50/sec; matched via veo-2
veo-3.0-generate-001	Google – Veo 3	API	$0.20/sec (720p/1080p); matched via veo-3
veo-3.0-fast-generate-001	Google – Veo 3 Fast	API	$0.08/sec (720p); matched via veo-3-fast
veo-3.1-generate-001	Google – Veo 3.1	API	$0.20/sec (720p/1080p); matched via veo-3.1
veo-3.1-fast-generate-001	Google – Veo 3.1 Fast	API	$0.08/sec (720p); matched via veo-3.1-fast
veo-3.1-lite-generate-001	Google – Veo 3.1 Lite	API	$0.03/sec (720p); matched via veo-3.1-lite
gemini-embedding-001	Google – Embedding	API	$0.00015/1K tokens
text-embedding-005	Google – Embedding	API	$0.000025/1K chars (per_thousand_tokens)
text-multilingual-embedding-002	Google – Embedding	API	$0.000025/1K chars (per_thousand_tokens)
textembedding-gecko@003	Google – Embedding	API	$0.000025/1K chars; legacy model
textembedding-gecko-multilingual@001	Google – Embedding	API	$0.000025/1K chars; legacy multilingual
text-embedding-large-exp-03-07	Google – Embedding	API	No dedicated row; using text-embedding pricing $0.000025/1K chars
gemini-embedding-2-preview	Google – Embedding	API	Gemini Embedding 2 (Unified Multimodal); $0.20/1M tokens
multimodalembedding@001	Google – Embedding	API	Multimodal embedding; $0.0002/1K chars text
claude-opus-4-6	Anthropic – Claude	API	@default stripped; input $5/$25; cache_write (5m) $6.25
claude-sonnet-4-6	Anthropic – Claude	API	@default stripped; input $3/$15; cache_write (5m) $3.75
claude-opus-4-5@20251101	Anthropic – Claude	API	Pinned date version; input $5/$25
claude-sonnet-4-5@20250929	Anthropic – Claude	API	Pinned date version; input $3/$15; long-context (>200K) $6/$22.50 noted
claude-haiku-4-5@20251001	Anthropic – Claude	API	Pinned date version; input $1/$5
claude-opus-4-1@20250805	Anthropic – Claude	API	Pinned date version; input $15/$75
claude-opus-4@20250514	Anthropic – Claude	API	Pinned date version; input $15/$75
claude-sonnet-4@20250514	Anthropic – Claude	API	Pinned date version; input $3/$15
gpt-oss-120b-maas	OpenAI	API	Matched "gpt-oss-120b" on page; $0.09/$0.36
llama-3.3-70b-instruct-maas	Meta – Llama	API	Matched "Llama 3.3 70B"; $0.72/$0.72
llama-4-maverick-17b-128e-instruct-maas	Meta – Llama 4	API	Matched "Llama 4 Maverick"; $0.35/$1.15
mistral-small-2503	Mistral AI	API	Matched "Mistral Small 3.1 (25.03)"; $0.10/$0.30
mistral-medium-3	Mistral AI	API	Matched "Mistral Medium 3"; $0.40/$2.00
codestral-2	Mistral AI	API	Matched "Codestral 2"; $0.30/$0.90
deepseek-r1-0528-maas	DeepSeek	API	Matched "DeepSeek-R1 (0528)"; $1.35/$5.40
deepseek-v3.1-maas	DeepSeek	API	Matched "DeepSeek-V3.1"; $0.60/$1.70; cache_read $0.06
deepseek-v3.2-maas	DeepSeek	API	Matched "DeepSeek-V3.2"; $0.56/$1.68; cache_read $0.056
qwen3-235b-a22b-instruct-2507-maas	Qwen	API	Matched "Qwen3 235B A22B (2507)"; $0.22/$0.88
qwen3-coder-480b-a35b-instruct-maas	Qwen	API	Matched "Qwen3 Coder 480B"; $0.22/$1.80; cache_read $0.022
qwen3-next-80b-a3b-instruct-maas	Qwen	API	Matched "Qwen3-Next 80B"; $0.15/$1.20
qwen3-next-80b-a3b-thinking-maas	Qwen	API	Matched "Qwen3-Next 80B" thinking variant; $0.15/$1.20
kimi-k2-thinking-maas	Moonshot AI – Kimi	API	Matched "Kimi-K2-Thinking"; $0.60/$2.50; cache_read $0.06
minimax-m2-maas	MiniMax	API	Matched "MiniMax-M2"; $0.30/$1.20; cache_read $0.03
glm-4.7-maas	ZAI.org – GLM	API	Matched "GLM-4.7"; $0.60/$2.20
glm-5-maas	ZAI.org – GLM	API	Matched "GLM-5"; $1.00/$3.20; cache_read $0.10
jamba-large-1.6	AI21	API – price not found	Self-deploy only (has_deploy: true, no -maas); no MaaS pricing row

Excluded Models

Model ID	Publisher	Reason
gemini-*-live-*	Google	Gemini Live streaming — separate product
lyria-*	Google	Music generation — no inference endpoint
model-optimizer-*	Google	Dynamic routing meta-endpoint
imagegeneration	Google	Legacy, superseded by imagen-3.0+
virtual-try-on-*	Google	Retail product model
gemma*, paligemma*, codegemma*	Google	Excluded per google.md (Gemma/non-generative)
chirp*	Google	Audio transcription, not generative inference
gemini-2.5-pro-tts / gemini-2.5-flash-tts	Google	Included with price 0 (no pricing row found)
clip-vit-base-patch32, openclip	OpenAI	Non-generative (feature extraction/classification)
whisper-large	OpenAI	Audio transcription
gpt-oss	OpenAI	Self-deploy (has_deploy: true, no -maas)
faster-r-cnn, retinanet, mask-r-cnn, segment-anything, sam3	Meta	Non-generative CV models
roberta-large, xlm-roberta-large	Meta	Non-generative NLP (self-deploy)
llama-guard, prompt-guard	Meta	Guard/safety models
llama2, llama3, llama3_1, llama3-2, llama3-3, llama4, codellama-7b-hf, llama-2-quantized, imagebind, nllb	Meta	Self-deploy (has_deploy: true, no -maas)
mistral, mixtral	Mistral AI	Self-deploy (mistral-ai publisher, has_deploy: true)
codestral-2501-self-deploy	Mistral AI	Self-deploy in name
mistral-ocr-2505	Mistral AI	OCR model
ministral-3, mistral-large-3	Mistral AI	Self-deploy (has_deploy: true, no -maas)
deepseek-r1, deepseek-v3, deepseek-v3-1, deepseek-v3-2	DeepSeek	Self-deploy (has_deploy: true, no -maas)
deepseek-ocr, deepseek-ocr-2, deepseek-ocr-maas	DeepSeek	OCR models
kimi-k2, kimi-k2-5	Moonshot AI	Self-deploy (has_deploy: true, no -maas)
minimax-m2	MiniMax	Self-deploy (has_deploy: true, no -maas)
glm-4.7, glm-5, glm-4.5	ZAI.org	Self-deploy (has_deploy: true, no -maas)
glm-ocr	ZAI.org	OCR model
glm-image	ZAI.org	Explicit policy exclusion (image gen excluded from Vertex AI pricing)
qwen-image	Qwen	Explicit policy exclusion (image gen excluded from Vertex AI pricing)
Remaining Qwen self-deploy models	Qwen	Self-deploy (has_deploy: true, no -maas)

Web Search Pricing by Generation

• Gemini 2.0 + 2.5: web_search $35/1000 → 3.5¢/search; enterprise_web_search $45/1000 → 4.5¢/search
• Gemini 3.x: web_search $14/1000 → 1.4¢/search; enterprise_web_search $14/1000 → 1.4¢/search

---

Generated by Pricing Agent on 2026-04-10