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)	11

➕ New Models

• gemini-2.5-pro-tts
• gemini-2.5-flash-tts
• claude-sonnet-4-6@default

🔄 Updated Models

• gemini-2.5-pro
• gemini-3-flash-preview
• gemini-3.1-pro-preview
• gemini-3.1-flash-lite-preview
• gemini-3.1-flash-image-preview
• gemini-3-pro-image-preview
• veo-3.0-fast-generate-preview
• veo-3.1-fast-generate-001
• text-embedding-005
• text-multilingual-embedding-002
• multimodalembedding

Model-to-Pricing-Page Mapping

Model ID	Publisher / Section	Source	Notes
gemini-2.5-pro	Google – Gemini 2.5 Pro	API	Standard (≤200K) input price used; tiered pricing >200K exists
gemini-2.5-flash	Google – Gemini 2.5 Flash	API	Standard input price
gemini-2.5-flash-lite	Google – Gemini 2.5 Flash-Lite	API	—
gemini-2.5-computer-use-preview-10-2025	Google – Gemini 2.5 Pro Computer Use	API	Maps to Gemini 2.5 Pro Computer Use pricing row
gemini-2.5-flash-preview-09-2025	Google – Gemini 2.5 Flash	API	Preview alias; matched to Gemini 2.5 Flash row
gemini-2.5-flash-lite-preview-09-2025	Google – Gemini 2.5 Flash-Lite	API	Preview alias; matched to Flash-Lite row
gemini-2.5-flash-image	Google – Gemini 2.5 Flash Image	API	Token pricing + image_token additional_unit
gemini-2.5-pro-tts	Google – Gemini 2.5 Pro	API – price not found	TTS model; no dedicated pricing row found
gemini-2.5-flash-tts	Google – Gemini 2.5 Flash	API – price not found	TTS model; no dedicated pricing row found
gemini-2.0-flash-001	Google – Gemini 2.0 Flash	API	—
gemini-2.0-flash-lite-001	Google – Gemini 2.0 Flash-Lite	API	—
gemini-3-pro-preview	Google – Gemini 3 Pro	API	Preview; matched to Gemini 3 Pro row
gemini-3-flash-preview	Google – Gemini 3 Flash	API	Preview; matched to Gemini 3 Flash row
gemini-3.1-pro-preview	Google – Gemini 3.1 Pro	API	Preview; matched to Gemini 3.1 Pro Preview row
gemini-3.1-flash-lite-preview	Google – Gemini 3.1 Flash-Lite	API	Preview; matched to Gemini 3.1 Flash-Lite row
gemini-3.1-flash-image-preview	Google – Gemini 3.1 Flash Image	API	Preview; matched to Gemini 3.1 Flash Image row; image_token additional
gemini-3-pro-image-preview	Google – Gemini 3 Pro Image	API	Preview; matched to Gemini 3 Pro Image row; image_token additional
imagen-3.0-generate-002	Google – Imagen 3	API	Per-image pricing $0.04/image
imagen-4.0-generate-001	Google – Imagen 4	API	Per-image pricing $0.04/image
imagen-4.0-fast-generate-001	Google – Imagen 4 Fast	API	Per-image pricing $0.02/image
imagen-4.0-ultra-generate-001	Google – Imagen 4 Ultra	API	Per-image pricing $0.06/image
imagen-3.0-capability-001	Google – Imagen 3 (capability)	API	Capability model; uses equivalent Imagen 3 generate price $0.04/image
imagen-3.0-capability-002	Google – Imagen 3 (capability)	API	Capability model; uses equivalent Imagen 3 generate price $0.04/image
veo-2.0-generate-001	Google – Veo 2	API	Video-only $0.50/s; duration 8s, sample 1
veo-3.0-generate-001	Google – Veo 3	API	Video-only 720p/1080p $0.20/s; duration 8s, sample 1
veo-3.0-fast-generate-001	Google – Veo 3 Fast	API	Video-only $0.10/s; duration 8s, sample 1
veo-3.0-generate-preview	Google – Veo 3	API	Preview; same pricing as Veo 3
veo-3.0-fast-generate-preview	Google – Veo 3 Fast	API	Preview; same pricing as Veo 3 Fast
veo-3.1-generate-001	Google – Veo 3.1	API	Video-only 720p/1080p $0.20/s; duration 8s, sample 1
veo-3.1-fast-generate-001	Google – Veo 3.1 Fast	API	Video-only $0.10/s; duration 8s, sample 1
veo-3.1-generate-preview	Google – Veo 3.1	API	Preview; same pricing as Veo 3.1
veo-3.1-fast-generate-preview	Google – Veo 3.1 Fast	API	Preview; same pricing as Veo 3.1 Fast
textembedding-gecko	Google – Text Embedding (legacy)	API – price not found	Legacy embedding; no pricing row found
text-embedding-005	Google – Text Embedding	API	$/1K chars ($0.000025/1K chars)
text-multilingual-embedding-002	Google – Text Multilingual Embedding	API	$/1K chars ($0.000025/1K chars)
text-embedding-large-exp-03-07	Google – Text Embedding Large (Exp)	API – price not found	Experimental; no dedicated pricing row
gemini-embedding-001	Google – Gemini Embedding	API	$0.00015/1K tokens online
gemini-embedding-2-preview	Google – Gemini Embedding 2	API – price not found	Preview; no dedicated pricing row
multimodalembedding	Google – Multimodal Embedding	API	text $0.0002/1K chars; image $0.0001/img; video plus/standard/essential per-second
claude-opus-4-6	Anthropic – Claude	API	API returns claude-opus-4-6@default; @default stripped
claude-sonnet-4-6	Anthropic – Claude	API	API returns claude-sonnet-4-6@default; @default stripped
claude-opus-4-5@20251101	Anthropic – Claude	API	Pinned @Date kept; matches Claude Opus 4.5 pricing row
claude-haiku-4-5@20251001	Anthropic – Claude	API	Pinned @Date kept; matches Claude Haiku 4.5 pricing row
claude-sonnet-4-5@20250929	Anthropic – Claude	API	Pinned @Date kept; matches Claude Sonnet 4.5 pricing row
claude-opus-4-1@20250805	Anthropic – Claude	API	Pinned @Date kept; matches Claude Opus 4.1 pricing row
claude-sonnet-4@20250514	Anthropic – Claude	API	Pinned @Date kept; matches Claude Sonnet 4 pricing row
claude-opus-4@20250514	Anthropic – Claude	API	Pinned @Date kept; matches Claude Opus 4 pricing row (no cache listed)
gpt-oss-120b-maas	OpenAI – GPT OSS	API	Matches gpt-oss-120b row; $0.09/$0.36 with batch
llama-3.3-70b-instruct-maas	Meta – Llama	API	Matches Llama 3.3 70B row; $0.72/$0.72 with batch
llama-4-maverick-17b-128e-instruct-maas	Meta – Llama 4 Maverick	API	Matches Llama 4 Maverick row; $0.35/$1.15 with batch
mistral-small-2503	Mistral AI – Small	API	Matches Mistral Small 3.1 (25.03) row; $0.10/$0.30
mistral-medium-3	Mistral AI – Medium	API	Matches Mistral Medium 3 row; $0.40/$2.00
codestral-2	Mistral AI – Codestral	API	Matches Codestral 2 row; $0.30/$0.90
deepseek-r1-0528-maas	DeepSeek – R1	API	Matches DeepSeek-R1 (0528) row; $1.35/$5.40 with batch
deepseek-v3.1-maas	DeepSeek – V3.1	API	Matches DeepSeek-V3.1 row; $0.60/$1.70 with cache and batch
deepseek-v3.2-maas	DeepSeek – V3.2	API	Matches DeepSeek-V3.2 row; $0.56/$1.68 with cache and batch
qwen3-235b-a22b-instruct-2507-maas	Qwen – Qwen3 235B	API	Matches Qwen3-235B-A22B-Instruct-2507 row; $0.22/$0.88 with batch
qwen3-coder-480b-a35b-instruct-maas	Qwen – Qwen3 Coder	API	Matches Qwen3-Coder-480B row; $0.22/$1.80 with cache and batch
qwen3-next-80b-a3b-instruct-maas	Qwen – Qwen3 Next 80B Instruct	API	Matches Qwen3-Next-80B-Instruct row; $0.15/$1.20
qwen3-next-80b-a3b-thinking-maas	Qwen – Qwen3 Next 80B Thinking	API	Matches Qwen3-Next-80B-Thinking row; $0.15/$1.20
minimax-m2-maas	MiniMax – M2	API	Matches MiniMax-M2 row; $0.30/$1.20 with cache
kimi-k2-thinking-maas	MoonshotAI – Kimi K2	API	Matches Kimi-K2-Thinking row; $0.60/$2.50 with cache
glm-4.7-maas	ZAI.org – GLM	API	Matches GLM-4.7 row; $0.60/$2.20
glm-5-maas	ZAI.org – GLM	API	Matches GLM-5 row; $1.00/$3.20 with cache

Excluded Models (not added)

Model	Publisher	Reason
imagegeneration	Google	Legacy; superseded by Imagen 3+
virtual-try-on-001	Google	Product-specific retail model
gemini-live-2.5-flash-native-audio	Google	*-live-* streaming model
lyria-002, lyria-3-pro-preview, lyria-3-clip-preview	Google	Music generation; no inference endpoint
All gemma*, paligemma, codegemma, etc.	Google	Self-deploy only (no -maas suffix)
All non-generative CV/NLP models (efficientnet, BERT, ResNet, etc.)	Google	Non-generative ML
pretrained-ocr	Google	OCR model
chirp-2, chirp-3	Google	Audio models — not generative text inference
translate-llm, text-translation	Google	Non-generative translation
video-text-detection, video-speech-transcription	Google	Non-generative video/audio analysis
weathernext, weather-next-v2	Google	Forecasting model — not generative AI inference
shieldgemma2	Google	Safety/guard model
earth-ai-imagery-*	Google	Specialized Earth observation models
clip-vit-base-patch32, openclip	OpenAI	Non-generative (CLIP embedding)
whisper-large	OpenAI	Audio transcription (not generative)
gpt-oss	OpenAI	Self-deploy only (no -maas)
segment-anything, sam3, faster-r-cnn, retinanet, mask-r-cnn	Meta	Non-generative CV (segmentation/detection)
xlm-roberta-large, roberta-large	Meta	Non-generative NLP
imagebind	Meta	Multimodal embedding (self-deploy)
nllb	Meta	Non-generative translation
llama-guard, prompt-guard	Meta	Guard/safety models
llama2, llama3, llama3_1, llama3-2, llama3-3, llama4, llama-2-quantized, codellama-7b-hf	Meta	Self-deploy only (no -maas)
mistral, mixtral	Mistral-AI (self-deploy publisher)	Self-deploy only
codestral-2501-self-deploy	Mistralai	Self-deploy only
mistral-ocr-2505	Mistralai	OCR model
ministral-3, mistral-large-3	Mistralai	Self-deploy only
deepseek-r1, deepseek-v3, deepseek-v3-1, deepseek-v3-2	DeepSeek-AI	Self-deploy only
deepseek-ocr, deepseek-ocr-2, deepseek-ocr-maas	DeepSeek-AI	OCR models
qwq, qwen3, qwen3-5, qwen2, qwen3-coder, qwen3-coder-next, qwen3-next, qwen3-embedding, qwen3-vl	Qwen	Self-deploy only
qwen-image	Qwen	Excluded by policy (image generation)
jamba-large-1.6	AI21	Self-deploy only
kimi-k2, kimi-k2-5	MoonshotAI	Self-deploy only
minimax-m2	MiniMax	Self-deploy only
glm-4.7, glm-5, glm-4.5	ZAI.org	Self-deploy only
glm-ocr	ZAI.org	OCR model
glm-image	ZAI.org	Excluded by policy (image generation)

---

Generated by Pricing Agent on 2026-03-31