Best AI for research
Updated 2026-10-16 from official prices and dated leaderboards.
Change budget or country in the finder
Up to US$30 a month (USA)
- Best pick Google AI Plus Google
US$4.99/mo
Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). - Also strong ChatGPT Plus OpenAI
US$20.00/mo
GPT-6 Astra is #2 on LiveBench 2.0 Reasoning (92.7). - Also consider Claude Pro Anthropic
US$20.00/mo · Annual billing US$17.00/mo
Claude Opus 5.5 is #4 on LiveBench 2.0 Reasoning (92.2).
Any budget (USA)
- Best pick Google AI Plus Google
US$4.99/mo
Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). - Also strong ChatGPT Plus OpenAI
US$20.00/mo
GPT-6 Astra is #2 on LiveBench 2.0 Reasoning (92.7). - Also consider Claude Pro Anthropic
US$20.00/mo · Annual billing US$17.00/mo
Claude Opus 5.5 is #4 on LiveBench 2.0 Reasoning (92.2).
In Canada, up to CA$30
Plans with an official CAD price come first; plans that only publish a US price follow, clearly labelled. We never convert.
- Best pick Google AI Plus Google
CA$6.99/mo
Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). - Also strong Claude Pro Anthropic
CA$28.00/mo · Annual billing CA$280.00/yr
Claude Opus 5.5 is #4 on LiveBench 2.0 Reasoning (92.2). - Also consider ChatGPT Free OpenAI
Free
GPT-5.6 Terra is #11 on LiveBench 2.0 Reasoning (90.6).
Who leads LiveBench 2.0 Reasoning
The leaderboard this page uses. Friendly model names, with the leaderboard's own name in grey. All rankings.
Higher is better. Bars compare rows in this table only.
| Rank | Model | Company | Score | Available in |
|---|---|---|---|---|
| #1 | Gemini 4 Argon | 95.1 | Google AI Plus, Google AI Pro, Google AI Ultra 5x, Google AI Ultra 20x | |
| #2 | GPT-6 Astra | OpenAI | 92.7 | ChatGPT Plus, ChatGPT Pro 100, ChatGPT Pro 200, ChatGPT Pro 500 |
| #3 | GPT-6.1 Sol | OpenAI | 92.6 | ChatGPT Plus, ChatGPT Pro 100, ChatGPT Pro 200, ChatGPT Pro 500 |
| #4 | Claude Opus 5.5 | Anthropic | 92.2 | Claude Pro, Claude Max 5x, Claude Max 20x |
| #5 | Claude Fable 5.1 | Anthropic | 91.7 | Claude Max 5x, Claude Max 20x |
| #6 | GPT-5.6 Sol (older version or setting) | OpenAI | 91.7 | – |
| #7 | Claude Sonnet 5.5 | Anthropic | 91.6 | Claude Free, Claude Pro, Claude Max 5x, Claude Max 20x |
| #8 | Claude Opus 5 (older version or setting) | Anthropic | 91.2 | – |
| #9 | Kimi K3 | Moonshot | 90.7 | We don't track this company's consumer plans |
| #10 | Muse Wonder 1.0 | Meta | 90.6 | Not named on plan pages |
| #11 | GPT-5.6 Terra | OpenAI | 90.6 | ChatGPT Free |
| #12 | Grok 4.6 | Grok (xAI) | 90.5 | Not named on plan pages |
Source: livebench.ai, checked 2026-10-15.
Every plan that fits (USA)
| Plan | USD / month | CAD / month | Why it's here |
|---|---|---|---|
| Google AI Plus Google | US$4.99 | CA$6.99 | Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). |
| Google AI Pro Google | US$19.99 | CA$26.99 | Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). |
| Google AI Ultra 5x Google | US$99.99 | CA$139.99 | Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). |
| Google AI Ultra 20x Google | US$199.99 | CA$279.99 | Gemini 4 Argon is #1 on LiveBench 2.0 Reasoning (95.1). |
| ChatGPT Plus OpenAI | US$20.00 | No official CAD price | GPT-6 Astra is #2 on LiveBench 2.0 Reasoning (92.7). |
| ChatGPT Pro 100 OpenAI | US$100.00 | No official CAD price | GPT-6 Astra is #2 on LiveBench 2.0 Reasoning (92.7). |
| ChatGPT Pro 200 OpenAI | US$200.00 | No official CAD price | GPT-6 Astra is #2 on LiveBench 2.0 Reasoning (92.7). |
| ChatGPT Pro 500 OpenAI | US$500.00 | No official CAD price | GPT-6 Astra is #2 on LiveBench 2.0 Reasoning (92.7). |
| ChatGPT Go OpenAI | US$8.00 | No official CAD price | OpenAI's best model on LiveBench 2.0 Reasoning is GPT-6 Astra (#2); the provider doesn't say which plans include it. |
| Claude Pro Anthropic | US$20.00 | CA$28.00 | Claude Opus 5.5 is #4 on LiveBench 2.0 Reasoning (92.2). |
| Claude Max 5x Anthropic | US$100.00 | CA$140.00 | Claude Opus 5.5 is #4 on LiveBench 2.0 Reasoning (92.2). |
| Claude Max 20x Anthropic | US$200.00 | CA$280.00 | Claude Opus 5.5 is #4 on LiveBench 2.0 Reasoning (92.2). |
| Claude Free Anthropic | Free | Free | Claude Sonnet 5.5 is #7 on LiveBench 2.0 Reasoning (91.6). |
| ChatGPT Free OpenAI | Free | Free | GPT-5.6 Terra is #11 on LiveBench 2.0 Reasoning (90.6). |
| X Subscription Plus Grok (xAI) | US$8.00 | No official CAD price | Grok (xAI)'s best model on LiveBench 2.0 Reasoning is Grok 4.6 (#12); the provider doesn't say which plans include it. |
| X Subscription Premium Grok (xAI) | US$30.00 | No official CAD price | Grok (xAI)'s best model on LiveBench 2.0 Reasoning is Grok 4.6 (#12); the provider doesn't say which plans include it. |
| X Subscription Super Grok (xAI) | US$100.00 | No official CAD price | Grok (xAI)'s best model on LiveBench 2.0 Reasoning is Grok 4.6 (#12); the provider doesn't say which plans include it. |
| X Subscription Ultra Grok (xAI) | US$200.00 | No official CAD price | Grok (xAI)'s best model on LiveBench 2.0 Reasoning is Grok 4.6 (#12); the provider doesn't say which plans include it. |
| Gemini Free Google | Free | Free | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
| X Subscription Free Grok (xAI) | Free | No official CAD price | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
| Perplexity Free Perplexity | Free | Free | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
| Perplexity Pro Perplexity | US$17.00 (annual billing) | No official CAD price | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
| Microsoft 365 Premium Microsoft | US$19.99 | CA$27.00 | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
| Microsoft 365 Pro Microsoft | US$99.99 | CA$135.00 | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
| Perplexity Max Perplexity | US$167.00 (annual billing) | No official CAD price | Includes what this goal needs, within budget. No ranked model is named for this plan, so it sorts by price. |
How we rank
We keep plans that include what this goal needs (needs web search or deep research), then sort by where the plan's best included model ranks on LiveBench 2.0 Reasoning, and break ties by price. One plan per provider in the picks. Nothing is blended. Where a provider doesn't say which plans include a ranked model, we say so. Prices come only from providers' own websites. Methodology.