Compare AI models

As of September 2026, Claude Pro charges extra for Claude Fable 5.1. llmwise Pro ($20) includes up to 31 Fable messages a month, counted before you send.

As of September 2026, ChatGPT Plus has GPT-6 Astra only in Work and Codex. llmwise has it in the chat: up to 31 messages a month on Pro ($20).

Chat with Kimi K3 in English, with no Kimi account. llmwise asks OpenRouter for hosts that don't store or train on prompts.

Z.ai's GLM 5.3 and GLM 5.3 Flash in English, no Z.ai account: in llmwise's free trial, then up to 250 GLM 5.3 a month or 60 Flash a day on Pro.

Chat with DeepSeek V4 Pro and V4.1 Flash in llmwise. Our routing asks OpenRouter for hosts that don't store or train on prompts.

Claude vs Gemini: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our test runs on 10 jobs, job by job.

Claude Haiku vs ChatGPT: Claude Haiku 4.5 gets up to 250 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs on 10 jobs, job by job.

DeepSeek vs Claude: DeepSeek V4 Pro gets up to 250 messages a month on llmwise Pro, Claude Sonnet 5.5 up to 125. Our test runs on 10 jobs, job by job.

DeepSeek vs Gemini: DeepSeek V4 Pro gets up to 250 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our test runs on 10 jobs, job by job.

DeepSeek vs ChatGPT: DeepSeek V4 Pro gets up to 250 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs on 10 jobs, job by job.

Claude vs ChatGPT: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs on 10 jobs, job by job.

ChatGPT vs Gemini: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our test runs on 10 jobs, job by job.

Grok vs Claude: Grok 4.7 gets up to 250 messages a month on llmwise Pro, Claude Sonnet 5.5 up to 125. Our test runs on 10 jobs, job by job.

Grok vs ChatGPT: Grok 4.7 gets up to 250 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs on 10 jobs, job by job.

ChatGPT vs Claude vs Gemini: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Claude Sonnet 5.5 up to 125, Gemini 3.1 Pro up to 125.

Gemini vs Grok: Gemini 3.1 Pro gets up to 125 messages a month on llmwise Pro, Grok 4.7 up to 250. Our test runs on 10 jobs, job by job.

Claude vs Kimi: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, Kimi K3 up to 125. Our test runs on 10 jobs, job by job.

DeepSeek vs Kimi: DeepSeek V4 Pro gets up to 250 messages a month on llmwise Pro, Kimi K3 up to 125. Our test runs on 10 jobs, job by job.

ChatGPT vs Claude vs Gemini vs Grok: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Claude Sonnet 5.5 up to 125, Gemini 3.1 Pro 125, Grok 4.7 250.

Claude vs GLM: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, GLM 5.3 up to 250. Our test runs on 10 jobs, job by job.

Kimi vs GLM: Kimi K3 gets up to 125 messages a month on llmwise Pro, GLM 5.3 up to 250. Our test runs on 10 jobs, job by job.

ChatGPT vs Kimi: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Kimi K3 up to 125. Our test runs on 10 jobs, job by job.

DeepSeek vs GLM: DeepSeek V4 Pro gets up to 250 messages a month on llmwise Pro, GLM 5.3 up to 250. Our test runs on 10 jobs, job by job.

Gemini vs GLM: Gemini 3.1 Pro gets up to 125 messages a month on llmwise Pro, GLM 5.3 up to 250. Our test runs on 10 jobs, job by job.

Grok vs GLM: Grok 4.7 gets up to 250 messages a month on llmwise Pro, GLM 5.3 up to 250. Our test runs on 10 jobs, job by job.

ChatGPT vs GLM: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, GLM 5.3 up to 250. Our test runs on 10 jobs, job by job.

Gemini vs Kimi: Gemini 3.1 Pro gets up to 125 messages a month on llmwise Pro, Kimi K3 up to 125. Our test runs on 10 jobs, job by job.

Grok vs DeepSeek: Grok 4.7 gets up to 250 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our test runs on 10 jobs, job by job.

Grok vs Kimi: Grok 4.7 gets up to 250 messages a month on llmwise Pro, Kimi K3 up to 125. Our test runs on 10 jobs, job by job.

AI models head to head, model by model: messages on every plan, the price of a message, context and files, plus the Claude, GPT and Gemini family pages.

Claude vs DeepSeek for coding: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our test runs, side by side.

Claude vs Gemini for coding: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our test runs, side by side.

DeepSeek vs ChatGPT for coding: DeepSeek V4 Pro gets up to 250 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs, side by side.

Claude vs ChatGPT for coding: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs, side by side.

ChatGPT vs Gemini for coding: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our coding test runs, reply by reply.

Grok vs Claude for coding: Grok 4.7 gets up to 250 messages a month on llmwise Pro, Claude Sonnet 5.5 up to 125. Our coding test runs, reply by reply.

Claude vs ChatGPT for writing: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs, side by side.

ChatGPT vs Gemini for writing: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our writing test runs, reply by reply.

Claude vs Gemini for writing: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our test runs, side by side.

ChatGPT vs Grok for coding: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Grok 4.7 up to 250. Our coding test runs, reply by reply.

ChatGPT vs Grok for writing: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Grok 4.7 up to 250. Our writing test runs, reply by reply.

ChatGPT vs DeepSeek for writing: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our test runs, side by side.

Claude vs DeepSeek for writing: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our test runs, side by side.

Claude vs ChatGPT for math: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our math test runs, reply by reply.

ChatGPT vs Gemini for math: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our math test runs, reply by reply.

Claude vs Gemini for math: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our math test runs, reply by reply.

ChatGPT vs Grok for math: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Grok 4.7 up to 250. Our math test runs, reply by reply.

ChatGPT vs DeepSeek for math: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our math test runs, reply by reply.

Claude vs DeepSeek for math: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our test runs, side by side.

Claude vs ChatGPT for data analysis: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, GPT-6.1 Sol up to 125. Our test runs, side by side.

ChatGPT vs Gemini for data analysis: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125. Our test runs, side by side.

Claude vs Gemini for data analysis: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, Gemini 3.1 Pro up to 125.

ChatGPT vs Grok for data analysis: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, Grok 4.7 up to 250. Our test runs, side by side.

ChatGPT vs DeepSeek for data analysis: GPT-6.1 Sol gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250. Our test runs, side by side.

Claude vs DeepSeek for data analysis: Claude Sonnet 5.5 gets up to 125 messages a month on llmwise Pro, DeepSeek V4 Pro up to 250.

The best AI chatbots, family by family: 19 models from 8 AI companies, tested on the same prompts, what each is for, and what a message costs.

The best AI for customer support, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for agents and tool use, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for coding, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for math, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for RAG and answering from documents, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for SQL, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for summarization, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for translation, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

The best AI for writing, from our test runs of all 19 models in llmwise: our picks, every reply, and what each costs per message.

19 AI models ranked by our own scored tests (50 prompts, 10 jobs), and 9 AI assistant apps' prices and free tiers in their own words.

The best AI for research, from our own scored tests: 19 models on 15 prompts that answer from documents, summarize and analyze data.

The best AI for students, from our own scored tests: 19 models on 20 math, writing, summary and translation prompts, and which are free.

The 16 models in llmwise's free trial, ranked by our own scored tests, and what 9 AI apps give free, in their own words.

Which AI model is best right now? All 19 models in llmwise ranked by our own scored tests: overall, on the hard prompts and job by job.

The best AI for business, from our own scored tests: 19 models on 20 writing, customer reply, data and summary prompts, and what each costs.

The best AI for data analysis, from our own scored tests: 19 models on 5 table questions with a checked answer, from an average to a correlation.

Which Claude model for coding (6 models, Fable 5.1 to Haiku 5.5), what each costs per message in llmwise, what matters, and tips.

Which Claude model for data analysis (6 models, Fable 5.1 to Haiku 5.5), what each costs per message in llmwise, what matters, and tips.

Which Claude model for writing (6 models, Fable 5.1 to Haiku 5.5), what each costs per message in llmwise, what matters, and tips.

Which DeepSeek model for coding (DeepSeek V4 Pro and DeepSeek V4.1 Flash), what each costs per message in llmwise, what matters, and tips.

Which DeepSeek model for customer support (DeepSeek V4 Pro and DeepSeek V4.1 Flash), what each costs per message in llmwise, what matters, and tips.

Which DeepSeek model for math (DeepSeek V4 Pro and DeepSeek V4.1 Flash), what each costs per message in llmwise, what matters, and tips.

Which DeepSeek model for summarization (DeepSeek V4 Pro and DeepSeek V4.1 Flash), what each costs per message in llmwise, what matters, and tips.

Which DeepSeek model for writing (DeepSeek V4 Pro and DeepSeek V4.1 Flash), what each costs per message in llmwise, what matters, and tips.

Which Gemini model for coding (Gemini 3.1 Pro (preview) and Gemini 3.8 Flash), what each costs per message in llmwise, what matters, and tips.

Which Gemini model for customer support (Gemini 3.1 Pro (preview) and Gemini 3.8 Flash), what each costs per message in llmwise, what matters, and tips.

Which Gemini model for data analysis (Gemini 3.1 Pro (preview) and Gemini 3.8 Flash), what each costs per message in llmwise, what matters, and tips.

Which Gemini model for math (Gemini 3.1 Pro (preview) and Gemini 3.8 Flash), what each costs per message in llmwise, what matters, and tips.

Which Gemini model for summarization (Gemini 3.1 Pro (preview) and Gemini 3.8 Flash), what each costs per message in llmwise, what matters, and tips.

Which Gemini model for writing (Gemini 3.1 Pro (preview) and Gemini 3.8 Flash), what each costs per message in llmwise, what matters, and tips.

Which GPT-6 model for coding (GPT-6 Astra, GPT-6.1 Sol, GPT-6 Sol, and GPT-6 Luna), what each costs per message in llmwise, what matters, and tips.

Which GPT-6 model for data analysis (GPT-6 Astra, GPT-6.1 Sol, GPT-6 Sol, and GPT-6 Luna), what each costs per message in llmwise, what matters, and tips.

Which GPT-6 model for math (GPT-6 Astra, GPT-6.1 Sol, GPT-6 Sol, and GPT-6 Luna), what each costs per message in llmwise, what matters, and tips.

Which GPT-6 model for summarization (GPT-6 Astra, GPT-6.1 Sol, GPT-6 Sol, and GPT-6 Luna), what each costs per message in llmwise, what matters, and tips.

Which GPT-6 model for writing (GPT-6 Astra, GPT-6.1 Sol, GPT-6 Sol, and GPT-6 Luna), what each costs per message in llmwise, what matters, and tips.

Which Grok model for coding (Grok 4.7), what each costs per message in llmwise, what matters, and tips.

Try Claude in your browser: Sonnet 5.5, Sonnet 5, Haiku 4.5, and Haiku 5.5 in the free trial, Claude Fable 5.1 on paid plans, next to 7 other AI families.

Chat with 19 AI models from 8 companies in one place: start with 5 free messages, see each model's count before you send, and pick by our test results.

Try Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral in one browser chat: free to start, with what each message costs shown before you send.

How many messages or images each llmwise plan gives on every model, what a long chat does to that, and what a $10 top-up adds.

Tick the AI subscriptions you pay for: see their total a month and a year, and what one llmwise plan gives on the same companies' models.

A model picker: choose the job, your plan and what the model must read, and see which LLMs in llmwise fit, by price band, with what each band suits.

Claude vs Llama: llmwise doesn't offer Llama, but has 6 Claude models. How the two differ in kind, when each makes sense, and the closest models llmwise has.

ChatGPT vs Llama: llmwise doesn't offer Llama, but has 4 GPT models. How the two differ in kind, when each makes sense, and the closest models llmwise has.