Skip to content

Models

The AI models behind chat apps and assistants: what each one costs, how much it can read at once, and whether the Vercel AI Gateway can keep your prompts from being stored or trained on.

All models26

Newest release first.

Source: Vercel AI Gateway, checked between 2026-09-25 and 2026-09-26. Prices in USD per 1M tokens.

how to read the table

What a model takes and returns

  • Text
  • Image
  • PDF
  • Audio
  • Video

reads as: takes text and images , then returns text

ZDR

Zero data retention, available on a paid Vercel plan through the AI Gateway (you turn it on).

NT

No training on your prompts, available through the gateway (you turn it on).

Marked only when every provider on the gateway offers it. “Some providers” means it depends on which provider serves you.

Models, newest release first
ModelContextInputOutputModalitiesDataReleased
Claude Opus 5.5By AnthropicContext1MInput /1M$4Output /1M$20Modalities: Takes , returns DataZDRNTReleased
GPT-6 LunaBy OpenAIContext1.05MInput /1M$0.10Output /1M$0.50Modalities: Takes , returns DataNTReleased
GPT-6 SolBy OpenAIContext1.05MInput /1M$2Output /1M$10Modalities: Takes , returns DataNTReleased
Grok 4.7By xAIContext500KInput /1M$1.20Output /1M$3.60Modalities: Takes , returns DataZDRNTReleased
MiMo V2.6 ProBy XiaomiContext1.05MInput /1M$0.435Output /1M$0.87Modalities: Takes , returns DataSome providersReleased
JevBy TypeSafeContext32KInput /1M$0.042Output /1M$0Modalities: Takes , returns DataNTReleased
DeepSeek V4.1 FlashBy DeepSeekContext1.05MInput /1M$0.30Output /1M$1.20Modalities: Takes , returns DataSome providersReleased
GPT-6 AstraBy OpenAIContext1.05MInput /1M$10Output /1M$50Modalities: Takes , returns DataNTReleased
Gemini 3.8 FlashBy GoogleContext1MInput /1M$0.75Output /1M$3.75Modalities: Takes , returns DataNTReleased
Claude Fable 5.1By AnthropicContext1MInput /1M$10Output /1M$50Modalities: Takes , returns DataNTReleased
Hunyuan Hy4 (preview)By TencentContext1.02MInput /1M$0.834Output /1M$2.501Modalities: Takes , returns DataNot offeredReleased
GLM-5.3By Z.aiContext1MInput /1M$1.40Output /1M$4.40Modalities: Takes , returns DataSome providersReleased
Qwen3.8 27BBy Alibaba CloudContext1MInput /1M$0.50Output /1M$3Modalities: Takes , returns DataSome providersReleased
Qwen 3.8 MaxBy Alibaba CloudContext262KInput /1M$2Output /1M$6Modalities: Takes , returns DataZDRNTReleased
Kimi K3By Moonshot AIContext1MInput /1M$3Output /1M$15Modalities: Takes , returns DataSome providersReleased
Hunyuan Hy3By TencentContext262KInput /1M$0.14Output /1M$0.58Modalities: Takes , returns DataSome providersReleased
Claude Sonnet 5By AnthropicContext1MInput /1M$2Output /1M$10Modalities: Takes , returns DataZDRNTReleased
Nemotron 3 UltraBy NVIDIAContext1MInput /1M$0.60Output /1M$2.40Modalities: Takes , returns DataZDRNTReleased
MiniMax M3By MiniMaxContext512KInput /1M$0.30Output /1M$1.20Modalities: Takes , returns DataSome providersReleased
Nano Banana 2By GoogleContext131KInput /1M$0.50Output /1M$3Modalities: Takes , returns DataNTReleased
DeepSeek V4 ProBy DeepSeekContext1MInput /1M$0.66Output /1M$1.98Modalities: Takes , returns DataSome providersReleased
Gemma 4 31BBy GoogleContext262KInput /1M$0.14Output /1M$0.40Modalities: Takes , returns DataSome providersReleased
Gemini 3.1 Pro (preview)By GoogleContext1MInput /1M$2Output /1M$12Modalities: Takes , returns DataNTReleased
Mistral Large 3By Mistral AIContext262KInput /1M$0.50Output /1M$1.50Modalities: Takes , returns DataZDRNTReleased
Sonar ProBy PerplexityContext200KInput /1MNot listedOutput /1MNot listedModalities: Takes , returns DataNTReleased
Muse Spark 1.3By MetaContext1.05MInput /1M$1.25Output /1M$4.25Modalities: Takes , returns DataZDRNTReleased Not listed

Friday

The brief, by email

Five to seven links an issue, each with what to do about it.

Unsubscribe from any issue.

This page as Markdown