ChatGPT, Gemini, Claude and Grok: How the Leading AI Models Differ

ChatGPT, Gemini, Claude and Grok differ in price, access and what each is tuned for. No single model leads every published test. As of October 6, 2026, the newest flagship models are:

• OpenAI's GPT-6 Astra
• Anthropic's Claude Fable 5.1 and Claude Opus 5.5
• Google's Gemini 3.8 Flash and limited-release Gemini 4 Argon
• xAI's Grok 4.7

Each chatbot runs on a family of underlying models that its maker updates several times a year.

Which AI model versions are current in October 2026?

The lineups below come from each lab's own documentation, checked on October 6, 2026.

OpenAI (ChatGPT). GPT-6 Astra was released on OpenAI's developer platform on September 3. It sits at the top of OpenAI's model list, which describes it as the company's most capable model. Cheaper options followed: GPT-6 Luna on September 22 and GPT-6.1 Sol on September 29.

Anthropic (Claude). Anthropic's lineup has four tiers. Claude Fable 5.1 is for the most demanding reasoning. Then come Claude Opus 5.5, Claude Sonnet 5.5 and the smaller Claude Haiku 4.5. Anthropic recommends Opus 5.5 as the starting point for most work.

Google (Gemini). Gemini 3.8 Flash, released on September 2, is the newest general-purpose Gemini model with a stable release. Google's developer documentation lists Gemini 3.1 Pro as a preview. On September 30 Google announced Gemini 4 Argon, which it describes as a frontier model. Google says Argon is going first to a group of trusted cybersecurity defenders, with paid API customers and Google AI Ultra subscribers next.

xAI (Grok). Grok 4.7 was released on September 21. The company, whose website now uses the name SpaceXAI, describes it as its most capable model for coding and knowledge work.

Anthropic offers Claude Mythos 5.1 by invitation only, and Google has given no date for Argon's wider release.

ChatGPT vs Gemini vs Claude vs Grok: what do they cost?

Developers pay by the token, a small chunk of text. Anthropic says one million tokens is roughly 555,000 words on its current models. List prices per million tokens, input then output, span a factor of 100:

• OpenAI: GPT-6 Astra, $10 and $50. GPT-6.1 Sol, $2 and $10. GPT-6 Luna, $0.10 and $0.50.
• Anthropic: Claude Fable 5.1, $10 and $50. Claude Opus 5.5, $4 and $20. Claude Sonnet 5.5, $2 and $10. Claude Haiku 4.5, $1 and $5.
• Google: Gemini 3.8 Flash, $0.75 and $3.75, an introductory rate that Google says rises to $1.50 and $7.50 on January 1, 2027. Gemini 4 Argon, announced at $2 and $10 for an introductory period and $4 and $20 afterward.
• xAI: Grok 4.7, from $2 and $6.

GPT-6 Astra and Claude Fable 5.1 carry identical list prices. One step down, OpenAI describes GPT-6.1 Sol as close to Astra's performance at lower cost. Anthropic says Opus 5.5 matches Fable 5.1 on most work.

Consumer subscriptions are priced separately. Anthropic lists Claude Pro at $20 a month when billed monthly. Google lists Google AI Pro at $19.99 a month.

Which AI model is best for coding?

It depends on the test, and most published scores come from the labs themselves. A comparison table that Google DeepMind published for Gemini 4 Argon shows three different leaders across four coding benchmarks. Argon is ahead on Vibe Code Bench and on DeepSWE v1.1. On DeepSWE v1.1 it scores 77.9%, against 74.2% for Claude Opus 5.5 and 74.1% for GPT-6 Astra. GPT-6 Astra leads FrontierSWE v2 with 65.5%, and Claude Opus 5.5 leads Terminal-Bench 4.0 with 66.4%. Grok 4.7 is not in the table.

Crowd-sourced rankings look different. The Arena text leaderboard ranks models by human votes. Its October 2, 2026 update put Gemini 4 Argon first, with a preliminary score of 1525. Five Anthropic entries followed, scoring between 1501 and 1505. The entry for GPT-6 Astra scored 1477, in 29th place. The entry for Grok 4.7 scored 1442, in 91st.

Neither settles the question. Stanford's 2026 AI Index says concerns about the reliability and gaming of widely used benchmarks are growing. It cites research suggesting an Arena ranking may partly reflect how well a model is adapted to that platform.

Which AI chatbot is the most popular?

By website visits, ChatGPT was still the largest by a wide margin in May 2026, but its lead had narrowed. Estimates from Similarweb, a web analytics company, show its share of worldwide visits to generative AI websites falling. By those estimates, the share went from about 76% in June 2025 to roughly 53% by May 2026. Its total visits held roughly steady over that time, the estimates show. Over the same period Gemini rose from under 9% to around 27% to 28%. Claude rose from about 2% to close to 9%. The page gives no comparable share for Grok.

The figures cover website visits only, not use inside mobile apps.

The bottom line

The evidence does not support naming a single winner. Lab-published coding benchmarks show different leaders on different tests. The October 2 Arena update put the newest flagship entries from the four companies as much as 83 points apart. Google has not said when Gemini 4 Argon will be widely available.