Three chatbots, three companies, one question on every signup screen: which $20 plan is worth it. ChatGPT Plus is $20/month, Claude Pro is $20/month ($17 billed annually), and Google AI Pro is $19.99/month. Same headline price, different message limits, different models, different defaults on whether your chats train the model. Here is the full matrix.
Claude vs ChatGPT vs Gemini: The Short Answer
All three run a free tier and a paid tier near $20/month. These differences change your day.
- ChatGPT has the widest ladder of plans (Free, Go $8, Plus $20, Pro at $100, $200, or $500). Plus runs GPT-6.1 Sol and GPT-6 Luna, and Codex is included on every plan, Free too.
- Claude bundles Claude Code into Pro at $20/month ($17 billed annually), runs a 1M-token context on all four current models, and lets you turn off model training to keep a 30-day data retention window instead of up to 5 years.
- Gemini is cheapest to upgrade ($4.99 AI Plus), puts Gemini 4 Argon on AI Pro, and includes the most cloud storage (5 TB on AI Pro).
If you write code in a terminal, Claude Opus 5.5 leads Terminal-Bench 4.0 even on Google's own launch table. If you want the cheapest paid step up, Gemini AI Plus is $4.99. If you want the longest single output, Gemini 4 Argon can write up to 1M tokens in one response.
Every Plan and Price
| Tier | ChatGPT | Claude | Gemini |
|---|---|---|---|
| Free | $0 (ads, US) | $0 | $0, 15 GB storage |
| Entry paid | Go: $8/mo | n/a | AI Plus: $4.99/mo |
| Standard paid | Plus: $20/mo | Pro: $20/mo ($17 annual) | AI Pro: $19.99/mo |
| Power tier | Pro: $100/mo | Max 5x: from $100/mo | AI Ultra: $99.99/mo (5x AI Pro) |
| Top tier | Pro: $200 or $500/mo | Max 20x | AI Ultra: $199.99/mo (20x AI Pro) |
| Cloud storage | n/a | n/a | Free 15 GB / Plus 400 GB / Pro 5 TB / Ultra 20 TB+ |
ChatGPT Go is $8/month and runs GPT-6 Luna. ChatGPT Pro comes at $100, $200, or $500 per month, and only the $500 plan gets Ultrafast mode in Codex. Claude Max starts at $100 for 5x Pro usage or 20x at the higher tier, and both Max tiers include $100 to $200 a month in Claude Platform API credits. Google AI Ultra is $99.99 for 5x AI Pro limits or $199.99 for 20x.
ChatGPT Business is $20/user/month annual ($25 monthly), minimum 2 users, with SAML SSO and admin controls. Claude Team runs 5 to 150 people at $20/seat annual ($25 monthly) for Standard, $100/seat for Premium. Claude Enterprise lists $20/seat with usage scaling by model, plus SCIM, audit logs, and HIPAA-ready controls.
Message and Usage Limits
Headline price tells you little about how much you can use each app. The limits decide it.
| Plan | ChatGPT | Claude | Gemini |
|---|---|---|---|
| Free | GPT-6 Luna, usage subject to rollout | Session + weekly limits | Flash-Lite model, base limits |
| Standard paid | Plus: GPT-6.1 Sol + GPT-6 Luna | Weekly limit (all models) + 5-hour session | AI Pro: 4x Free usage limits, Gemini 4 Argon |
| How usage is pooled | ChatGPT Work and Codex share one pool, extendable with credits | Claude Code counts against the same limits | AI Plus gets 2x Free limits |
| Power tier | Pro: $100, $200, or $500 tiers | Max 5x or 20x Pro usage | Ultra: 5x or 20x AI Pro usage |
None of the three publishes a fixed message count in October 2026. OpenAI says ChatGPT Work and Codex share usage, so agent runs and chats draw from one pool, and paid plans can extend it with ChatGPT credits. Claude uses a 5-hour session limit plus a weekly limit across all models. Google states its tiers as multiples of Free, with AI Plus at 2x and AI Pro at 4x, and AI Ultra at 5x or 20x AI Pro.
Which Model Each App Runs
| ChatGPT | Claude | Gemini | |
|---|---|---|---|
| Top model | GPT-6 Astra | Claude Fable 5.1 | Gemini 4 Argon (Oct 13, 2026) |
| $20-tier models | GPT-6.1 Sol, GPT-6 Luna | More Claude models than Free, Claude Code | Gemini 4 Argon, Pro, Flash |
| Free-tier model | GPT-6 Luna | Fewer models than Pro | Flash-Lite |
| Context window | 1.05M tokens (API) | 1M tokens | 1M tokens |
| Max output | 128K tokens | 128K tokens | Up to 1M tokens |
| Knowledge cutoff | Apr 30, 2026 (Astra, Sol) | Jun 2026 | not published here |
ChatGPT Plus runs GPT-6.1 Sol and GPT-6 Luna. OpenAI added GPT-6.1 Terra on October 26, 2026 at $1/$5 per million API tokens, and GPT-5.5 left Codex for ChatGPT sign-ins on October 14. Anthropic lists Fable 5.1, Opus 5.5, Sonnet 5.5, and Haiku 5.5 as current, each with a 1M-token context, 128K max output, and a June 2026 knowledge cutoff. Google released Gemini 4 Argon to the Gemini app on October 13 for AI Pro and Ultra subscribers. On the API it costs $2/$10 per million tokens through November 30, 2026, then $4/$20, and it is not on the free tier.
Latest GPT, Claude, and Gemini Model Releases
Each vendor shipped a new model between October 26 and November 10, 2026. The table lists the API release date, model ID, and list price per million tokens.
| Released | Model | Vendor | API price per 1M tokens (in / out) |
|---|---|---|---|
| Nov 10, 2026 | Gemini 4 Carbon (gemini-4-carbon) | $2 / $10 through Nov 30 | |
| Nov 9, 2026 | Claude Fable 5.2 (claude-fable-5-2) | Anthropic | $8 / $40 |
| Oct 26, 2026 | GPT-6.1 Terra (gpt-6.1-terra) | OpenAI | $1 / $5 |
| Oct 13, 2026 | Gemini 4 Argon (gemini-4-argon) | $2 / $10 through Nov 30, then $4 / $20 |
Claude Fable 5.2 costs less than Fable 5.1 ($8 / $40 per million tokens, down from $10 / $50) and keeps the 1M-token context and 128K max output. Google calls Gemini 4 Carbon its most capable reasoning model. Argon uses fewer tokens per turn than Carbon. On the OpenAI API, GPT-6.1 Sol lists at $2 / $10 and GPT-6 Astra at $10 / $50. The plan and app tables above reflect each vendor's app pages as checked on October 27, 2026.
ChatGPT vs Claude vs Gemini Benchmarks
Each vendor publishes its own numbers on its own scaffold, and the benchmark variants differ, so read these as directional rather than a head-to-head leaderboard. The scores come from Google's October 13 Gemini 4 Argon launch. Claude Fable 5.2 and Gemini 4 Carbon shipped after it and are not in this table.
| Benchmark | GPT-6 Astra | Claude Opus 5.5 | Gemini 4 Argon |
|---|---|---|---|
| Terminal-Bench 4.0 | 58.2% | 66.4% | 57.4% |
| DeepSWE v1.1 | 74.1% | 74.2% | 77.9% |
| FrontierSWE v2 | 65.5% | 62.3% | 55.0% |
| Vals Index | 63.1% | 67.0% | 68.9% |
| GraphWalks BFS, 256K to 1M (F1) | 71.8% | 66.8% | 84.2% |
Google ran all four models in the same table, Claude Fable 5.1 included (57.9% Terminal-Bench 4.0, 67.4% DeepSWE v1.1). Each coding benchmark has a different winner. Opus 5.5 leads terminal work, GPT-6 Astra leads FrontierSWE v2, and Argon leads DeepSWE v1.1 and long-context retrieval. Anthropic's own Opus 5.5 launch table puts Astra at 57.9% on Terminal-Bench 4.0, so expect a point or two of spread between vendor harnesses.
Does Your Data Train the Model
This is the setting most people never check, and the default differs across all three.
| ChatGPT | Claude | Gemini | |
|---|---|---|---|
| Trains on chats by default | Yes | User choice | Yes |
| How to opt out | Settings > Data Controls | Model improvement setting | Turn off Keep Activity |
| Retention if you opt out | History kept, not trained on | 30-day retention | Auto-delete (default 18 months) |
| Retention if you opt in | Used for training | De-identified up to 5 years | Activity older than 18 months auto-deleted |
ChatGPT's "Improve the model for everyone" toggle is on by default for consumer accounts; switching it off in Settings means new conversations are not used to train models, and the change applies account-wide and can be reversed anytime. Gemini trains by default via the Keep Activity setting; turning it off stops future chats from being used to improve Google AI (audio, Live videos, and screenshares are already excluded by default). Claude makes model improvement a user choice on Free, Pro, and Max: leave it off and Anthropic does not use your chats for training and applies a 30-day retention window; opt in and data may be retained de-identified for up to 5 years.
Pick by Use Case
| You want | Best pick | Why |
|---|---|---|
| Cheapest paid upgrade | Google AI Plus | $4.99/mo, 2x free limits, 400 GB. |
| Longest single output | Google AI Pro | Gemini 4 Argon writes up to 1M tokens per response. |
| Terminal coding | Claude Pro | Claude Code bundled at $20/mo; Opus 5.5 leads Terminal-Bench 4.0. |
| Privacy-first default | Claude | Model training is a choice; 30-day retention when off. |
| Long documents | Claude or Google AI Pro | 1M-token context on every current Claude model and on Argon. |
| Video generation | Gemini | Included from AI Plus up. |
| Google Workspace user | Gemini | Embedded across Gmail, Docs, Drive. |
Building on These Models Yourself
The consumer apps wrap a model in a chat box. If you are building a product on top of these models or running open-source models like DeepSeek, the chat subscription is the wrong unit; you want an API and per-token pricing. Morph serves open-source and coding-tuned models with full 16-bit activations (no fp8 quantization) so output matches the reference weights, plus codegen-specific speculative decoding for fast apply. See Morph Open Source Models and pricing for the per-token rates.
Frequently Asked Questions
Which is cheapest: ChatGPT, Claude, or Gemini?
All three have a free tier. The cheapest paid plan is Google AI Plus at $4.99/month, then ChatGPT Go at $8/month. The standard paid tiers are within a dollar: ChatGPT Plus $20, Claude Pro $20 ($17 annual), Google AI Pro $19.99.
How many messages do you get on the free plans?
None of the three publishes a fixed free message count. ChatGPT Free runs GPT-6 Luna. Gemini Free runs a Flash-Lite model, and AI Plus doubles its limits. Claude Free has session and weekly usage limits that vary by demand.
Which model does each chatbot run?
ChatGPT Plus runs GPT-6.1 Sol and GPT-6 Luna, with GPT-6.1 Terra added October 26, 2026. Claude runs Fable 5.1, Opus 5.5, Sonnet 5.5, and Haiku 5.5. Google AI Pro and Ultra run Gemini 4 Argon (October 13, 2026) with a 1M-token context; the free tier uses a Flash-Lite model.
Does ChatGPT, Claude, or Gemini train on my conversations?
ChatGPT and Gemini train by default (opt out in Settings > Data Controls and by turning off Keep Activity, respectively). Claude makes it a user choice; left off, it does not train on your chats and keeps a 30-day retention window.
What are the newest ChatGPT, Claude, and Gemini models?
Google released Gemini 4 Carbon on the Gemini API on November 10, 2026 at $2 / $10 per million tokens through November 30. Anthropic released Claude Fable 5.2 on November 9 at $8 / $40, down from $10 / $50 for Fable 5.1. OpenAI's newest API model is GPT-6.1 Terra (October 26) at $1 / $5.
Which is best for coding?
Each coding benchmark picks a different winner. On Google's launch table Claude Opus 5.5 leads Terminal-Bench 4.0 (66.4%) and is bundled with Claude Code. GPT-6 Astra leads FrontierSWE v2 (65.5%), and ChatGPT ships Codex. Gemini 4 Argon leads DeepSWE v1.1 (77.9%). For running open-source coding models on an API, see Morph Open Source Models.
Related comparisons
ChatGPT vs Gemini
GPT-5.5 vs Gemini 3.1 Pro: benchmarks, pricing, multimodal, and when to route to each.
ChatGPT vs Gemini vs Copilot
Three consumer assistants, three ecosystems: OpenAI, Google, and Microsoft 365.
ChatGPT vs Claude vs Grok
OpenAI, Anthropic, and xAI head to head on reasoning, coding, and tone.
Claude vs ChatGPT
An honest comparison from a team that routes production traffic to both.
ChatGPT vs Microsoft Copilot
Copilot runs on OpenAI models inside Microsoft 365. Where the wrapper helps and where it taxes you.
ChatGPT vs Grok
GPT-5.5 vs Grok: real-time X data and fewer guardrails vs ecosystem maturity.
Building on These Models?
The consumer apps are a chat box. If you are shipping a product or running open-source models like DeepSeek, Morph serves them on an API with full 16-bit activations and codegen-tuned speculative decoding.
