LLM Reference

Vercel AI Gateway Models — Pricing & Benchmarks

181 models available · Vercel

Vercel AI Gateway hosts 181 AI models in this catalog. The lowest listed input price is Amazon Titan Text Embeddings V2 at $0.02/1M input tokens. LLM Reference lets you compare these models across all 65 providers without switching tabs.

ModelInput (per 1M)Output (per 1M)Context
Amazon Titan Text Embeddings V2$0.02——
Mistral NeMo (2407)$0.02$0.04128K
text-embedding-3-small$0.02—8K
KAT Coder Pro V1$0.03$1.2256K
Amazon Nova Micro$0.035$0.144k
Trinity Mini$0.045$0.15128K
GPT-5 Nano$0.05$0.4400K
gpt-oss-20b$0.05$0.2131K
Nemotron 3 Nano 30B-A3B$0.05$0.24—
Amazon Nova Lite$0.06$0.24300k
Nemotron-Nano-9B-v2$0.06$0.23—
GLM-4.7 Flash$0.07$0.4198K
Gemini 2.0 Flash-Lite$0.075$0.31.048576M
GPT OSS Safeguard 20B$0.075$0.3131K
Qwen3-30B-A3B$0.08$0.29128K
Devstral Small 2$0.1$0.3256K
Gemini 2.5 Flash Lite$0.1$0.41M
GPT-4.1 Nano$0.1$0.41M
Llama 3.2 1B Instruct$0.1$0.1128K
Mistral Ministral 3B$0.1$0.1—
Mistral Small 3$0.1$0.333K
Qwen3.5-Flash$0.1$0.41M
text-embedding-ada-002$0.1—8K
Xiaomi MiMo-V2-Flash$0.1$0.3262K
Embed v4.0$0.12—128k
Qwen3-14B$0.12$0.2440K
Gemma 4 26B A4B IT$0.13$0.4256k
text-embedding-3-large$0.13—8K
DeepSeek V4 Flash$0.14$0.281M
Gemma 4 31B$0.14$0.4256k
Gemini 2.0 Flash$0.15$0.62M
Gemini Embedding$0.15——
GPT-4o-mini$0.15$0.6128K
GPT-4o-mini Search Preview$0.15$0.6128K
Llama 3.2 3B Instruct$0.15$0.15128K
Ministral 8B$0.15$0.1532k
Nemotron 3 Super-120B-A12B$0.15$0.651M
Pixtral 12B Instruct$0.15$0.15128K
Qwen3-Coder-30B-A3B-Instruct$0.15$0.6—
Llama 3.2 11B Vision Instruct$0.16$0.16128K
Qwen3-32B$0.16$0.6440K
Llama 4 Scout 17B Instruct$0.17$0.6610M
Gemini Embedding 2$0.2——
GLM-4.5-Air$0.2$1.1128K
GPT-5.4 Nano$0.2$1.25400K
Ministral 14B$0.2$0.232k
Nemotron-Nano-12B-v2-VL$0.2$0.6—
Llama 3.1 8B Instruct$0.22$0.22128K
Llama 4 Maverick 17B Instruct$0.24$0.971M
Claude 3 Haiku$0.25$1.25200K
Gemini 3.1 Flash Lite Preview$0.25$1.51M
Gemini 3.1 Flash-Lite$0.25$1.51M
GPT-5 Mini$0.25$2400K
GPT-5.1 Codex Mini$0.25$2400K
Mercury 2$0.25$0.75131K
Seed 1.6$0.25$2256K
Trinity-Large-Preview$0.25$1128K
Trinity-Large-Thinking$0.25$0.9256K
DeepSeek V3.1 Terminus$0.27$1164K
DeepSeek V3.2$0.28$0.42160K
Amazon Nova 2 Lite$0.3$2.51M
Gemini 2.5 Flash$0.3$2.51M
GLM 4.6V$0.3$0.9128K
KAT Coder Pro V2$0.3$1.2256K
MiniMax M2$0.3$1.2197K
MiniMax M2.1$0.3$1.2200k
MiniMax M2.5$0.3$1.2197K
MiniMax M2.7$0.3$1.2205K
Mistral Codestral 2508$0.3$0.9256K
Nano Banana (Gemini 2.5 Flash Image)$0.3$2.533K
gpt-oss-120b$0.35$0.75131K
GPT-4.1 Mini$0.4$1.61M
Qwen3 VL 235B A22B Instruct$0.4$1.6256K
Qwen3.5-Plus$0.4$2.41M
DeepSeek V4 Pro$0.435$0.871M
Gemini 3 Flash$0.5$31M
GPT-3.5 Turbo$0.5$1.516K
Mistral Large 3 675B Instruct$0.5$1.5128K
Mistral Magistral Small 2509$0.5$1.5—
Nano Banana 2 (Gemini 3.1 Flash Image Preview)$0.5$366K
Qwen3-Coder-Next$0.5$1.2256K
Qwen3.6-Plus$0.5$31M
DeepSeek V3.1$0.56$1.6864K
Kimi K2 Instruct$0.57$2.3131K
GLM 4.5V$0.6$1.864K
GLM-4.5$0.6$2.2128K
GLM-4.6$0.6$2.2198K
Kimi K2 Thinking$0.6$2.5256K
Kimi K2.5$0.6$3256K
MiniMax M2.5 Highspeed$0.6$2.4205K
MiniMax M2.7 Highspeed$0.6$2.4205K
Qwen3.6-27B$0.6$3.6262K
Llama 3.1 70B Instruct$0.72$0.72128K
Llama 3.2 90B Vision Instruct$0.72$0.72128K
Llama 3.3 70B Instruct (free)$0.72$0.7266K
GPT-5.4 Mini$0.75$4.5400K
DeepSeek V3$0.77$0.7764k
Amazon Nova Pro$0.8$3.2500k
Claude 3.5 Haiku$0.8$4200k
Morph V3 Fast$0.8$1.280K
Morph V3 Large$0.9$1.9256K
Kimi K2.6$0.95$4262K
Claude Haiku 4.5$1$5200k
GLM-5$1$3.2200k
Grok Build 0.1$1$2256K
MiMo-V2-Pro$1$31M
Qwen3-Coder-Plus$1$51M
o3 Mini$1.1$4.4200K
o4-mini$1.1$4.4200K
Kimi K2 Thinking Turbo$1.15$8262K
GLM-5 Turbo$1.2$4200k
GLM-5V-Turbo$1.2$4200k
Qwen3-Max$1.2$6128K
Gemini 2.5 Pro$1.25$101M
GPT-5$1.25$10400K
GPT-5 Chat$1.25$10128K
GPT-5 Codex$1.25$10400K
GPT-5.1 Codex$1.25$10400K
GPT-5.1 Codex Max$1.25$10—
Grok 4.20 Multi-Agent$1.25$2.52M
Grok 4.20 Non-Reasoning$1.25$2.51M
Grok 4.20 Reasoning$1.25$2.51M
Grok 4.3$1.25$2.51M
Qwen3.6 Max Preview$1.3$7.8256K
DeepSeek R1$1.35$5.4128K
GLM-5.1$1.4$4.4200k
Gemini 3.5 Flash$1.5$91M
GPT-3.5 Turbo (Instruct)$1.5$24K
Mistral Medium 3.5$1.5$7.5256K
Qwen3-Coder-480B-A35B-Instruct$1.5$7.5256K
GPT-5.2$1.75$14400K
GPT-5.2 Codex$1.75$14—
GPT-5.3 Chat$1.75$14128K
GPT-5.3-Codex$1.75$14400K
Gemini 3 Pro Preview$2$121M
Gemini 3.1 Pro Preview$2$121M
GPT Image 1 Mini$2$8—
GPT-4.1$2$81M
o3$2$8200K
Pixtral Large$2$6128K
GLM 4.7$2.25$2.75200K
Command A$2.5$10256k
GPT-4o$2.5$10128K
GPT-5.4$2.5$151.1M
Qwen3.7-Max$2.5$7.51M
Claude Sonnet 4$3$151M
Claude Sonnet 4.5$3$15200K
Claude Sonnet 4.6$3$151M
Claude Opus 4.5$5$25200K
Claude Opus 4.6$5$251M
Claude Opus 4.7$5$251M
GPT Image 1$5$40—
GPT Image 1.5$5$32—
GPT Image 2$5$30—
GPT-5.5$5$301.1M
GPT-4 Turbo$10$30128K
o3 Deep Research$10$40200K
Claude Opus 4.1$15$75200k
GPT-5 Pro$15$120400K
o3-pro$20$80200K
GPT-5.2 Pro$21$168400K
GPT-5.4 Pro$30$1801.1M
GPT-5.5 Pro$30$1801.1M
Cohere Rerank 3.5———
FLUX.1 Kontext [pro]———
FLUX.1.1 [pro]———
FLUX.1.1 [pro] Ultra———
Grok Imagine Image———
Grok Imagine Video———
Imagen 4———
Imagen 4 Fast———
Imagen 4 Ultra———
Recraft V3———
Recraft V4———
Seedream 4.5———
Sonar——127K
Sonar Pro——200K
Sonar Reasoning Pro——128K
Veo 3———
Veo 3 Fast———
Veo 3.1———

Pricing Overview

Cheapest$0.02/1M
Most expensive$30.00/1M

About Vercel AI Gateway

Vercel AI Gateway is a unified AI proxy providing a single OpenAI-compatible API endpoint to 275+ models from 25+ providers including Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, xAI, Alibaba, Amazon, ByteDance, Cohere, MiniMax, MoonshotAI, KwaiPilot, Black Forest Labs, Recraft, Voyage AI, NVIDIA, and more. Pricing is pass-through at provider list rates with zero markup. Includes $5/month free tier; paid is pay-as-you-go. Features: automatic provider fallbacks, unified observability, streaming, tool use, vision, embeddings, image/video generation, BYOK mode. Integrates via @ai-sdk/gateway package or plain model ID strings in Vercel AI SDK. API details: API key via Authorization: Bearer <AI_GATEWAY_API_KEY>. Key from Vercel dashboard. Free $5/month credit; paid tier is provider list price with zero markup. BYOK (bring-your-own-key) also supported with no markup or fee. Model IDs use {provider-owner}/{model-name} — e.g., anthropic/claude-opus-4.6, openai/gpt-5.

Full provider profile →

Links

Dashboard Documentation Pricing