Skip to content

AI data

Updated · By , AI-assisted.

Messages per $20: every AI plan that publishes a count, and llmwise's

On llmwise Pro ($20), $20 buys up to 31 GPT-6 Astra messages a month, up to 125 on Claude Sonnet 5, and 60 everyday messages a day. Of 13 AI subscriptions we track, none publishes a chat count, so the table sets llmwise beside the chat apps that publish one.

llmwise: messages per $20, model by model

Each model's count on each llmwise plan, scaled to $20 of the plan's price: Plus is $10, Pro is $20, Max is $50, Ultra is $100, Studio is $200 a month. Everyday models are a day's count on each plan.

Messages per $20 on each llmwise plan
ModelPlus ($10)Pro ($20)Max ($50)Ultra ($100)Studio ($200)
Claude Fable 5.1Not on Plus31 a month40 per $20 (100 a month)45 per $20 (225 a month)50 per $20 (500 a month)
Claude Opus 5.562 per $20 (31 a month)62 a month80 per $20 (200 a month)90 per $20 (450 a month)100 per $20 (1,000 a month)
Claude Sonnet 5.5124 per $20 (62 a month)125 a month160 per $20 (400 a month)180 per $20 (900 a month)200 per $20 (2,000 a month)
Claude Sonnet 5124 per $20 (62 a month)125 a month160 per $20 (400 a month)180 per $20 (900 a month)200 per $20 (2,000 a month)
Claude Haiku 5.530 a day60 a day120 a day200 a day200 a day
Claude Haiku 4.5250 per $20 (125 a month)250 a month320 per $20 (800 a month)360 per $20 (1,800 a month)400 per $20 (4,000 a month)
GPT-6 AstraNot on Plus31 a month40 per $20 (100 a month)45 per $20 (225 a month)50 per $20 (500 a month)
GPT-6.1 Sol124 per $20 (62 a month)125 a month160 per $20 (400 a month)180 per $20 (900 a month)200 per $20 (2,000 a month)
GPT-6 Sol124 per $20 (62 a month)125 a month160 per $20 (400 a month)180 per $20 (900 a month)200 per $20 (2,000 a month)
GPT-6 Luna30 a day60 a day120 a day200 a day200 a day
Gemini 4 Argon62 per $20 (31 a month)62 a month80 per $20 (200 a month)90 per $20 (450 a month)100 per $20 (1,000 a month)
Gemini 3.1 Pro (preview)124 per $20 (62 a month)125 a month160 per $20 (400 a month)180 per $20 (900 a month)200 per $20 (2,000 a month)
Gemini 3.8 Flash250 per $20 (125 a month)250 a month320 per $20 (800 a month)360 per $20 (1,800 a month)400 per $20 (4,000 a month)
DeepSeek V4.1 Flash30 a day60 a day120 a day200 a day200 a day
DeepSeek V4 Pro250 per $20 (125 a month)250 a month320 per $20 (800 a month)360 per $20 (1,800 a month)400 per $20 (4,000 a month)
Grok 4.7250 per $20 (125 a month)250 a month320 per $20 (800 a month)360 per $20 (1,800 a month)400 per $20 (4,000 a month)
Kimi K3124 per $20 (62 a month)125 a month160 per $20 (400 a month)180 per $20 (900 a month)200 per $20 (2,000 a month)
GLM 5.3250 per $20 (125 a month)250 a month320 per $20 (800 a month)360 per $20 (1,800 a month)400 per $20 (4,000 a month)
GLM 5.3 Flash30 a day60 a day120 a day200 a day200 a day
Mistral Large 4250 per $20 (125 a month)250 a month320 per $20 (800 a month)360 per $20 (1,800 a month)400 per $20 (4,000 a month)

Every paid plan also has a fair-use ceiling on monthly AI cost: $3.75 on Plus, $7.50 on Pro, $20 on Max, $42 on Ultra, $85 on Studio. The counts here assume typical message sizes, which stay under it; outsized messages and heavy research hit it first. Every limit, explained.

Other plans that publish a count

Only counts the company publishes, in its words. A weekly count is times 52 over 12; a yearly plan's price is its monthly equivalent.

Published counts per $20
PlanPriceWhat it publishesA monthPer $20
ChatHub Pro (yearly)$14.99 a month

“1500 advanced model queries / month” (ChatHub pricing)

1,500 on its advanced models2,001
Monica Pro (yearly)$8.25 a month

“200 advanced model queries monthly” (Monica pricing)

200 on its advanced models484
Writingmate Ultimate (yearly)$39.99 a month

“Top models — about 500 messages” (Writingmate pricing)

500 on its top models250
Writingmate Pro (yearly)$19.99 a month

“Top models — about 200 messages” (Writingmate pricing)

200 on its top models200

Plans that publish no count

These describe their limits in words, not a count, so there's nothing to divide. Here's what each says.

Plans without a published count
PlanFromIts words on limits
ChatGPT Go$8 a month

“ChatGPT Go includes higher limits for eligible features such as image generation, file uploads, and data analysis.” (OpenAI Help Center: What is ChatGPT Go?)

ChatGPT Plus$20 a month

“To ensure a smooth experience for all users, Plus subscriptions may include usage limits such as message caps, especially during high demand. These limits may vary based on system conditions.” (OpenAI Help Center: What is ChatGPT Plus?)

ChatGPT Pro$100 a month

“Pro 200 includes more usage than Pro 100. Pro 500 offers the highest included usage of the three plans.” (OpenAI Help Center: About ChatGPT Pro tiers)

Claude Pro$20 a month

“The number of messages you can send will vary based on message length, including the length of files you attach, the length of your current conversation, and the model or feature you use. Your session-based usage limit will reset every five hours.” (Claude Help Center: What is the Pro plan?)

Claude Max$100 a month

“Max 5x includes five times the Pro plan's per-session usage allowance.” (Claude Help Center: What is the Max plan?)

Google AI Pro$19.99 a month

“AI Pro: 4x higher than standard limits” (Gemini Apps Help: limits and upgrades for Google AI subscribers)

Google AI Ultra$99.99 a month

“AI Ultra: 5x or 20x higher than AI Pro limits depending on your subscription” (Gemini Apps Help: limits and upgrades for Google AI subscribers)

SuperGrok$30 a month

“Smarter answers in Expert mode: Access our best model with higher limits” (Grok plans)

Perplexity Pro$20 a month

“More usage limits and memory” (Perplexity pricing)

Poe$19.99 a month

“660 thousand points/month” (Poe subscription plans)

Copilot Pro$19.99 a month

“AI features only available to subscription owner and cannot be shared; usage limits apply.” (Microsoft Store: Microsoft 365 Premium)

Work out your own stack's total in the subscription stack calculator, or one model and plan in the message calculator.

More AI data

Bar chart: Messages on llmwise Pro ($20 a month), by model. Claude Haiku 4.5: up to 250 a month; Gemini 3.8 Flash: up to 250 a month; DeepSeek V4 Pro: up to 250 a month; Grok 4.7: up to 250 a month; GLM 5.3: up to 250 a month; Mistral Large 4: up to 250 a month; Claude Sonnet 5.5: up to 125 a month; Claude Sonnet 5: up to 125 a month; GPT-6.1 Sol: up to 125 a month; GPT-6 Sol: up to 125 a month; Gemini 3.1 Pro (preview): up to 125 a month; Kimi K3: up to 125 a month; Claude Opus 5.5: up to 62 a month; Gemini 4 Argon: up to 62 a month; Claude Fable 5.1: up to 31 a month; GPT-6 Astra: up to 31 a month.
llmwise's published counts. A monthly count assumes the whole allowance goes to that model; everyday models count by the day.

Messages per $20: how we counted

Why do most plans have no messages-per-$20 figure?

Because most don't publish one. ChatGPT Plus, Claude Pro, Google AI Pro, SuperGrok and Perplexity Pro describe their limits in words (multiples of a free tier, resets every few hours) rather than counts, so there's no number to divide. The table quotes what each says instead.

Is a bigger plan better value per message?

A bigger llmwise plan gives more per dollar: $20 of Studio buys 50 GPT-6 Astra messages to Pro's 31. Everyday models are a daily count on every plan, and don't draw on the month.

How is another plan's figure worked out?

A weekly count times 52 weeks over 12 months, then scaled to $20 of the plan's monthly price, rounded down. A count on a yearly plan uses its monthly equivalent. Only counts the company itself publishes are used.

Spend the $20 where the count is published

On llmwise Pro, $20 is up to 31 Claude Fable 5.1 messages a month or up to 250 Claude Haiku 4.5 messages a month, with 60 everyday messages a day on top.