Skip to content

New: DeepSeek V4.1 Flash Derisked. In chat and on the API.

See modelsWhat's newReport a bug
Open chat

Every model in Vani, side by side.

The same models in Vani Chat and on the Vani API. See what each is best for, how much it can read at once, whether it reads images and how quickly it answers — 35 models, 1 new.

The whole sky

35 models, 10 labs

Size and colour: API output price, $0.33 to $35 per 1M tokens
De-risked, 18+
Under each name: input / output per 1M tokens

35 of 35 models

By lab, as on the chart

A star's size and colour: API output price, $0.33 to $35 per 1M tokens.

Every language model in Vani Chat and on the Vani API, grouped by lab: API output price per 1M tokens, context window, whether it reads images, thinking effort, speed, what it is and the plans it is on
ModelAPI price: Output / 1MContextImagesThinkingSpeedWhat it is, and the plans it is on
Anthropic10 models
Claude Opus 5.5, Anthropic$14.00$14.00 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxStandardStandard

Claude Opus 5.5 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare Claude Opus 5.5Details of Claude Opus 5.5

Claude Sonnet 5.5, Anthropic$7.00$7.00 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

Claude Sonnet 5.5 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Writing · Coding · Long documentsPlus and aboveCompare Claude Sonnet 5.5Details of Claude Sonnet 5.5

Claude Haiku 5.5, Anthropic$0.35$0.35 output per 1M1M1M contextReads imagesReads imagesFixed thinking effortFastFast

Claude Haiku 5.5 is Anthropic's fast, light model with a 1M-token context window. It reads images, answers quickly and is light on your allowance.

Best for Everyday · Quick answers · Long documentsAll plansCompare Claude Haiku 5.5Details of Claude Haiku 5.5

Claude Fable 5.1, Anthropic$35.00$35.00 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxSpeed not measured yet

Claude Fable 5.1 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Hard problems · Creative writingPlus and aboveCompare Claude Fable 5.1Details of Claude Fable 5.1

Claude Fable 5, Anthropic$35.00$35.00 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxSpeed not measured yet

Claude Fable 5 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare Claude Fable 5Details of Claude Fable 5

Claude Opus 5, Anthropic$17.50$17.50 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxSpeed not measured yet

Claude Opus 5 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Hard problems · Analysis · CodingPlus and aboveCompare Claude Opus 5Details of Claude Opus 5

Claude Sonnet 5, Anthropic$7.00$7.00 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

Claude Sonnet 5 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Writing · CodingPlus and aboveCompare Claude Sonnet 5Details of Claude Sonnet 5

Claude Opus 4.8, Anthropic$17.50$17.50 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxSpeed not measured yet

Claude Opus 4.8 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare Claude Opus 4.8Details of Claude Opus 4.8

Claude Opus 4.7, Anthropic$17.50$17.50 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighSpeed not measured yet

Claude Opus 4.7 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare Claude Opus 4.7Details of Claude Opus 4.7

Claude Opus 4.6, Anthropic$17.50$17.50 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighSpeed not measured yet

Claude Opus 4.6 is an Anthropic model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare Claude Opus 4.6Details of Claude Opus 4.6

OpenAI8 models
GPT-6.1 Sol, OpenAI$7.00$7.00 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxDeliberateDeliberate

GPT-6.1 Sol is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Research · Study · CodingPlus and aboveCompare GPT-6.1 SolDetails of GPT-6.1 Sol

GPT-6.1 Terra, OpenAI$3.50$3.50 output per 1M1M1M contextReads imagesReads imagesExtra highThinking effort, up to Extra highStandardStandard

GPT-6.1 Terra is a balanced, cost-efficient model in OpenAI’s GPT-6.1 family, positioned below GPT-6.1 Sol. Designed for everyday agentic workflows, coding, and content creation, it excels at interpreting screenshots and visual interfaces while delivering reliable performance and efficiency for high-volume workloads.

Best for Everyday taks · automations · budget friendly.All plansCompare GPT-6.1 TerraDetails of GPT-6.1 Terra

GPT-6 Astra, OpenAI$35.00$35.00 output per 1M1M1M contextReads imagesReads imagesMaxThinking effort, up to MaxDeliberateDeliberate

GPT-6 Astra is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Hard problems · AnalysisPlus and aboveCompare GPT-6 AstraDetails of GPT-6 Astra

GPT-6 Sol, OpenAI$7.00$7.00 output per 1M1M1M contextReads imagesReads imagesExtra highThinking effort, up to Extra highSpeed not measured yet

GPT-6 Sol is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare GPT-6 SolDetails of GPT-6 Sol

GPT-6 LunaFree to try, OpenAI$0.35$0.35 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

GPT-6 Luna is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Everyday · Quick answersAll plansCompare GPT-6 LunaDetails of GPT-6 Luna

GPT-5.6 Sol, OpenAI$14.00$14.00 output per 1M1M1M contextReads imagesReads imagesExtra highThinking effort, up to Extra highSpeed not measured yet

GPT-5.6 Sol is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Research · CodingPlus and aboveCompare GPT-5.6 SolDetails of GPT-5.6 Sol

GPT-5.6 Terra, OpenAI$8.40$8.40 output per 1M1M1M contextReads imagesReads imagesExtra highThinking effort, up to Extra highSpeed not measured yet

GPT-5.6 Terra is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Plus and aboveCompare GPT-5.6 TerraDetails of GPT-5.6 Terra

GPT-5.6 Luna, OpenAI$0.84$0.84 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

GPT-5.6 Luna is an OpenAI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

All plansCompare GPT-5.6 LunaDetails of GPT-5.6 Luna

Google2 models
Gemini 3.8 Flash, Google$2.625$2.625 output per 1M1M1M contextReads imagesReads imagesFixed thinking effortSpeed not measured yet

Gemini 3.8 Flash is a Google model with a 1M-token context window. It reads images.

Plus and aboveCompare Gemini 3.8 FlashDetails of Gemini 3.8 Flash

Gemini 3.1 Pro, Google$8.40$8.40 output per 1M1M1M contextReads imagesReads imagesFixed thinking effortStandardStandard

Gemini 3.1 Pro is a Google model with a 1M-token context window. It reads images.

Plus and aboveCompare Gemini 3.1 ProDetails of Gemini 3.1 Pro

xAI3 models
Grok 4.7, xAI$4.20$4.20 output per 1M500K500K contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

Grok 4.7 is an xAI model with a 500K-token context window. It reads images and lets you choose how hard it thinks.

All plansCompare Grok 4.7Details of Grok 4.7

Grok 4.6, xAI$4.20$4.20 output per 1M500K500K contextReads imagesReads imagesHighThinking effort, up to HighDeliberateDeliberate

Grok 4.6 is an xAI model with a 500K-token context window. It reads images and lets you choose how hard it thinks.

Best for Everyday · Quick answersAll plansCompare Grok 4.6Details of Grok 4.6

Grok 4.5, xAI$4.20$4.20 output per 1M500K500K contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

Grok 4.5 is an xAI model with a 500K-token context window. It reads images and lets you choose how hard it thinks.

All plansCompare Grok 4.5Details of Grok 4.5

Z.AI3 models
GLM-5.3, Z.AI$3.08$3.08 output per 1M1M1M contextReads imagesReads imagesHighThinking effort, up to HighStandardStandard

GLM-5.3 is a Z.AI model with a 1M-token context window. It reads images and lets you choose how hard it thinks.

Best for Long documents · WritingPlus and aboveCompare GLM-5.3Details of GLM-5.3

GLM 5.3 Flash, Z.AI$0.35$0.35 output per 1M1M1M contextDoes not read imagesHighThinking effort, up to HighStandardStandard

GLM 5.3 Flash is a Z.AI model with a 1M-token context window. It lets you choose how hard it thinks.

All plansCompare GLM 5.3 FlashDetails of GLM 5.3 Flash

GLM 5.3 DeriskedDe-riskedFree to try, Z.AI$6.50$6.50 output per 1M1M1M contextDoes not read imagesHighThinking effort, up to HighStandardStandard

GLM 5.3 Derisked is a Z.AI model with a 1M-token context window. It lets you choose how hard it thinks.

Best for Fiction for adults · Frank answersAll plansCompare GLM 5.3 DeriskedDetails of GLM 5.3 Derisked

DeepSeek2 models
DeepSeek V4.1 Flash, DeepSeek$0.42$0.42 output per 1M1M1M contextDoes not read imagesFixed thinking effortStandardStandard

DeepSeek V4.1 Flash is a DeepSeek model with a 1M-token context window.

All plansCompare DeepSeek V4.1 FlashDetails of DeepSeek V4.1 Flash

DeepSeek V4.1 Flash DeriskedDe-riskedNew, DeepSeek$2.60$2.60 output per 1M1M1M contextDoes not read imagesMaxThinking effort, up to MaxStandardStandard

DeepSeek V4.1 Flash with far fewer refusals, and a 1M-token context window. For adults (18+).

Best for Fiction for adults · Frank answersAll plansCompare DeepSeek V4.1 Flash DeriskedDetails of DeepSeek V4.1 Flash Derisked

Moonshot AI4 models
Kimi K3.1 Pro, Moonshot AI$7.00$7.00 output per 1M1M1M contextDoes not read imagesExtra highThinking effort, up to Extra highStandardStandard

Kimi K3.1 Pro is a Moonshot AI model with a 1M-token context window. It lets you choose how hard it thinks.

Plus and aboveCompare Kimi K3.1 ProDetails of Kimi K3.1 Pro

Kimi K3.1 Pro Soft-DeriskedDe-risked, Moonshot AI$7.00$7.00 output per 1M1M1M contextDoes not read imagesExtra highThinking effort, up to Extra highDeliberateDeliberate

Fewer refusals than standard Kimi K3.1 Pro, more conservative than GLM 5.3 Derisked. It's Kimi K3.1 Pro with Vani's soft-derisked instructions. For adults (18+).

Best for Long-context research and analysis · CodingPlus and aboveCompare Kimi K3.1 Pro Soft-DeriskedDetails of Kimi K3.1 Pro Soft-Derisked

Kimi K3, Moonshot AI$10.50$10.50 output per 1M1M1M contextDoes not read imagesExtra highThinking effort, up to Extra highSpeed not measured yet

Kimi K3 is a Moonshot AI model with a 1M-token context window. It lets you choose how hard it thinks.

Best for Coding · AnalysisPlus and aboveCompare Kimi K3Details of Kimi K3

Kimi K2.8 Preview, Moonshot AI$2.80$2.80 output per 1M1M1M contextDoes not read imagesHighThinking effort, up to HighStandardStandard

Kimi K2.8 Preview is a Moonshot AI model with a 1M-token context window. It lets you choose how hard it thinks.

All plansCompare Kimi K2.8 PreviewDetails of Kimi K2.8 Preview

MiniMax1 model
MiniMax M3, MiniMax$0.84$0.84 output per 1M1M1M contextDoes not read imagesFixed thinking effortSpeed not measured yet

MiniMax M3 is a MiniMax model with a 1M-token context window.

All plansCompare MiniMax M3Details of MiniMax M3

Qwen1 model
Qwen3.8 Flash, Qwen$0.329$0.329 output per 1M1M1M contextDoes not read imagesFixed thinking effortStandardStandard

Qwen3.8 Flash is a Qwen model with a 1M-token context window.

All plansCompare Qwen3.8 FlashDetails of Qwen3.8 Flash

Tencent1 model
Hy4 Preview, Tencent$1.7507$1.7507 output per 1M1M1M contextDoes not read imagesFixed thinking effortStandardStandard

Hy4 Preview is a Tencent model with a 1M-token context window.

All plansCompare Hy4 PreviewDetails of Hy4 Preview

Intelligence rank: Based on data from Artificial Analysis (artificialanalysis.ai). The ranking is Vani's ordering of the models it offers and is not endorsed by Artificial Analysis. Source (opens in a new tab)

Image and video models

Make and edit images and video in Vani Chat and Vani Studio, or call the same models on the API. Prices are prepaid API prices per output; in chat and Studio, generations use your plan's allowance. Typical time is the median of recent successful generations.

Pick a model. Change it mid-conversation.

Every model here is in Vani Chat and on the Vani API. Guest chat needs no account.