Claude
Free runs Sonnet 5.5, near the top for coding and hard questions. Pay for Pro for agents and for writing with Opus 5.5.
This product is made by Anthropic, whose Claude models help draft our reviews. How we keep that from tilting a score
Should you use it?
Pick it if
People who want strong free help with coding and hard questions, or will pay $20 a month for Opus 5.5 for agent work and writing.
Best for: Free help with coding and hard questions; Pro for agent work and writing (AI apps Top 20)
Skip it if
You mainly write: free Sonnet 5.5 ranks only 24th of 412 on Creative Writing s4, so pay for Pro or Gemini AI Pro, both 70 for writing.
You give up: Free plan's model is weaker at creative writing; usage limits are not given in numbers
The short version
- What this is
- Anthropic's assistant app on web, desktop and mobile. Free runs Sonnet 5.5; Pro adds Opus 5.5, Claude Code and Cowork features.
- Why it matters
- Its free plan already runs a model near the top of the coding and hard-question boards, so many people need not pay.
- Free plan model is Sonnet 5.5, 3rd on arena.ai WebDev and 3rd on Text Expert s1 s2 s5 s6.
- Pro at $20/month adds Opus 5.5, which leads the Agent board s3 s5.
- You choose in privacy settings whether chats are used for training s8.
Good at
- The free plan's model, Sonnet 5.5, ranks 3rd of 142 on arena.ai WebDev and 3rd of 364 on Text Expert s1 s2 s5 s6.
- Pro adds Opus 5.5, first on the arena.ai Agent board and second on Creative Writing s3 s4 s5.
- Opus 5.5 won 4 of 5 everyday prompts against ChatGPT-6 in a Tom's Guide face-off; the ChatGPT-6 version was not named s10.
Weak at
- Anthropic does not publish usage limits in numbers; Max is sold as five times Pro usage every 5 hours and every week, and Ultra as ten times Pro's weekly usage with no 5-hour limit for verified accounts s5.
- Sonnet 5.5, the free plan's model, ranks only 24th of 412 on Creative Writing s4.
- Free has no Claude Code, Claude in Chrome or Research, so agent work needs Pro s5.
Our take
The case for Claude is its free plan. Anthropic's pricing page lists Sonnet and Haiku on Free without naming a default s5, so under our rule for every app, the best model a plan lets you choose, Free is scored on Sonnet, now Sonnet 5.5 s11; PCWorld reports it is the free default s6. Sonnet 5.5 sits 3rd on both the WebDev and Text Expert boards s1 s2, which earns Free a score of 80 for coding and hard questions at $0. Free limits are not published.
Pro at $20 a month adds Opus, so Pro is scored on its best model for each job: Opus 5.5, first on the Agent board and second on Creative Writing s3 s4. That gives Pro 70 for agents and writing. Fable is sold to Pro only as extra usage credits, so we did not count it; Max includes it, but Fable 5.1 ranks below Opus 5.5 or Sonnet 5.5 on all four boards s1 s2 s3 s4. The ranks are for high-effort variants, and Anthropic says its apps default to Medium effort s11, so treat each score as that plan's ceiling.
We checked for the best non-Anthropic option at each job: LM Studio also scores 80 for coding at $0, Perplexity Pro's 70 is the best for hard questions (ChatGPT's free plan scores 60), ChatGPT Plus also scores 70 for agents, Gemini AI Pro 70 for writing, and GPT-6.1 Sol is the pick through the API.
This product is made by Anthropic, whose Claude models help draft our reviews.
Scores by job
A score from 0 to 100 per job, against everything available today. Overall 75 across 4 jobs, the average of the job scores.
New to this? What the scores and labels mean
- 0 to 100
- Our score for one job, like writing or coding, compared with everything else out today. 100 is the best available today. For AI models we build it from public data, such as Arena's blind votes and Epoch AI's independent tests; apps and tools are scored on our own checklist. The same tool can score differently for different jobs, and where two likely ranges overlap, we call it a tie.
- Our verdict
- Use it: a top choice for most people. Worth trying: good, and worth a go. Depends: good for some people or one kind of job. Wait: too early to tell, or not out for everyone yet. Skip: something else does the job better.
- First look
- Just released. We say what it is and whether to bother, and give it a score once public results exist.
- No score yet
- None of the public sources we use covers it yet, or the only evidence is from a source whose numbers we can't use, so we describe it in words instead.
- Best buy
- Nothing cheaper scores as well for that job.
- Best free
- The best score for that job that costs nothing, through a free plan or by running on your own computer.
- Pay as you go
- No monthly plan: you pay for each use, usually through a developer account.
- Checked
- The day we last confirmed the score and the price against the sources. Prices change often, so check the date.
- AI model
- The engine underneath, like GPT or Claude. You use it through an app or, if you code, directly.
- Open source
- Free to download and run on your own computer, so nothing leaves your machine.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so far
Sources and notes
Public leaderboardMaker’s pagesFree (lists Sonnet and Haiku with no default named, so the better one, Sonnet, now Sonnet 5.5, s5, s11; PCWorld reports it is the free default, s6): 8/10, does the job 2 (arena.ai WebDev rank 3 of 142, claude-sonnet-5.5-xhigh, s1), underlying quality 2, value 2 ($0, s5), privacy 1 (training is a setting you choose, but the default is not stated, so the lower score, s8), limits 1 (Free limits not published). Pro (lists Opus, Sonnet and Haiku with no default named, so the best for coding, Opus 5.5, s5, s7): 7/10, does the job 2 (WebDev rank 1 of 142, claude-opus-5.5-max, s1), underlying quality 2, value 1 ($20/month, s5), privacy 1, limits 1. Score from Free, the higher total. Sources s1 s5 s7.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so far
Sources and notes
Public leaderboardMaker’s pagesFree (Sonnet 5.5, as above, s5, s11): 8/10, does the job 2 (arena.ai Text Expert rank 3 of 364, spread 1-31, claude-sonnet-5.5-xhigh, 512 votes, s2), underlying quality 2, value 2 ($0), privacy 1 s8, limits 1. Pro (best for research is also Sonnet 5.5, ahead of Opus 5.5 at rank 6, s2): 7/10, does the job 2, underlying quality 2, value 1 ($20/month), privacy 1, limits 1. Score from Free, the higher total.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so farOn the $20/month plan
Sources and notes
Public leaderboardMaker’s pagesFree: not scored; the pricing page lists Claude Code, Claude in Chrome, Research and task hand-off only from Pro s5. Pro (Opus 5.5, its best model for agents, s5, s7): 7/10, does the job 2 (Claude Opus 5.5 (High) arena.ai Agent rank 1 of 52, +14.33%, s3), underlying quality 2, value 1 ($20/month), privacy 1 s8, limits 1 (not published). Score from Pro.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so farOn the $20/month plan
Sources and notes
Public leaderboardMaker’s pagesFree (Sonnet 5.5, s5, s11): 6/10, does the job 1 (arena.ai Creative Writing rank 24 of 412, claude-sonnet-5.5-xhigh, s4), underlying quality 1, value 2 ($0), privacy 1 s8, limits 1. Pro (Opus 5.5, its best model for writing, s5, s7): 7/10, does the job 2 (rank 2 of 412, claude-opus-5.5-high, s4), underlying quality 2, value 1 ($20/month), privacy 1, limits 1. Score from Pro, the higher total.
Alternatives
Better or cheaper for coding, from our scores, then the ones we'd also look at. The same kind of product comes first.
- ChatGPT (GPT-6)App70$20 a monthScores 70 for agents and coding on Plus with GPT-6 Astra, at the same $20 price, with images and voice. Side by side
- Claude Opus 5.5BetterText model100$20 a monthScores 100 in coding, against 80 here; $20 a month. The model to use through the API if you would rather pay per task than per month. Side by side
- GPT-6.1 SolText model85$20 a monthBest non-Anthropic API pick we review: #4 on Arena's Agent board and 85 for coding (Arena WebDev, rank 4), at half Opus 5.5's token price. Side by side
Value for money
Where it sits in coding: higher is a better score, further left is cheaper. Green dots are the Best buys.
Evidence
Everything behind the call and the scores, folded so the page stays short. Spotted a wrong price, score or link? Report a mistake.
How we scored this
The coding, research, agents and writing scores come from our own checklist of the product.
Public leaderboardMaker’s pages
This product is made by Anthropic, whose Claude models help draft our reviews.
Reviews are drafted with help from AI models, including Anthropic's Claude. Model scores come only from outside leaderboards and independent evaluations, by a written rule.
No one paid for this review or gave us free access, and we earn nothing if you sign up. How we score
Sources 11
- s1arena.ai Code Arena: WebDev leaderboard
- s2arena.ai Text: Expert leaderboard
- s3arena.ai Agent leaderboard
- s4arena.ai Text: Creative Writing leaderboard
- s5Claude plans and pricing
- s6PCWorld: Claude Sonnet 5.5 is here (29 Sep 2026)
- s7Claude release notes
- s8Claude privacy: is my data used for model training (consumer)
- s9Anthropic: Updates to our consumer terms
- s10Tom's Guide: ChatGPT-6 vs Claude Opus 5.5 with 5 everyday prompts (24 Sep 2026)
- s11Anthropic: Introducing Claude Sonnet 5.5 (28 Sep 2026)
Something here out of date or wrong? Report a mistake and we will check it against the source.
History 7 changes
Every change to a rating, the price or our call, with the reason. Scores are against what exists on the day, so they drift down as better things ship. Older entries come from before we moved to scores.
- Plans: Anthropic renamed its top plans. Max ($100/month) is five times Pro usage every 5 hours and every week; Ultra ($200/month, formerly Max 20x) is ten times Pro's weekly usage with no 5-hour limit for verified accounts. Prices unchanged; no score changes.
- New: Coding ratedNew: Research ratedNew: Agents ratedNew: Writing ratedReviewed: coding, research, agents, writing.
- Audit: underlying-quality note updated after Sonnet 5.5 coding moved down a step; it is still in the top band, so the score is unchanged.
- Better: Writing upAudit under our rule that each plan is rated on the best model it includes: Anthropic names no default for Pro, so Pro is now rated on its best model for each job instead of an assumed Sonnet 5.5. Writing on Pro rose to 7/10 on Opus 5.5. Other ratings unchanged. Dropped the 27 Aug Tom's Guide test, which predates the current models, and stopped counting the Opus 5.5 test toward the Free plan, which runs Sonnet 5.5.
- Rule change: best model each plan lets you choose. Free still offers Sonnet and Haiku and Pro adds Opus, so no rating changes. Arena text URLs moved to the /leaderboard/chat/text/ form.
- Like-with-like check: ratings compare Free with Pro only. Max adds Fable, which ranks below Opus 5.5 or Sonnet 5.5 on every board we use, so no top-plan note changes a score. No rating changes.
- Marked the agents and writing ratings as needing Pro ($20/month), since both are rated from the Pro plan. Ratings unchanged.
Use it well 4 guides on this site
Pages on this site that help with the jobs above.