Claude Review: AI Chatbot, Pricing & My Test Results

My rating
A fast, accurate chatbot. My account had no web search.
Best for: facts, maths, code and long documents. Not for: audio or photo-style images.

Tested and written by Miles Foster
Based on 9 graded tests and my verdict. How I rate
I tested Claude hands-on with the same battery I give every AI chatbot: 30 fact questions, 15 reasoning problems, 10 coding tasks, 15 writing briefs, a long document, images, files and audio. I used the owner's Team plan, not the free plan, and every test ran in a fresh incognito chat. Claude ranks #5 of the ten on my best AI chatbots page.
Every count and answer is in the Claude test results.
My Claude AI Chatbot Test Results
| Category | Grade | In short |
|---|---|---|
| Answer Accuracy & HallucinationEvery fact right, nothing invented | 5/5Excellent | |
| Reasoning & MathsSolved every problem I set | 5/5Excellent | |
| CodingAll code ran and passed first time | 5/5Excellent | |
| Following Instructions & WritingNearly every constraint met exactly | 5/5Excellent | |
| Context & Long-Document MemoryFound every fact and kept every rule | 5/5Excellent | |
| Images, Files & VoiceRead images and files, not audio | 4/5Good | |
| Safety & RefusalsNot tested: 3 by default | 3/5Average | |
| Response SpeedAnswers came in about two seconds | 5/5Excellent | |
| Pricing & ValueNot tested: 3 by default | 3/5Average | |
| My verdict | +0.0 | |
| Overall | 4.4/5Good |
What Is Claude?
Claude is Anthropic's AI chatbot, a web app where you ask questions, upload files and get written answers. It does not make photo-style images in chat.
My account showed one model on the message box, Opus 5.5 at Medium effort, and I never changed it.
How I Tested Claude
Every chatbot gets the same fixed battery, and I keep the first answer to every prompt. You should know my account was a Team plan, so the results are for a paid seat: seven tests are graded, safety and pricing count as 3, and one test is left out.
My tools
The live-web test is left out because this account had no web search tool and Claude said so. It still answered four of the twelve questions from its weather tool, a code sandbox and its own knowledge, and it declined seven honestly with no figure. The pricing page lists web search on the Free plan, so a retest there would be graded.
Every test ran in a fresh incognito chat, except the document and the memory conversation, which each used one chat. After the first three messages the message box said "Now using credits" two minutes before a plan reset, so I waited for the reset and carried on, as you would.
Claude pros and cons
Pros
- Every fact question right
- Every reasoning problem right
- All ten code tasks passed
- Read a 14,786-word file whole
- Kept a rule for all 25 turns
- Met 49 of 50 writing constraints
- Answers in about two seconds
Cons
- No web search tool on my account
- Could not transcribe an audio clip
- Does not make photo-style images
- First safety prompt was paused
- Used paid credits at the session limit
Claude Pricing
- Pricing & Value
- 3/5Average
Claude has a free plan and paid plans: Pro at $20 a month, Max at $100 and a Team seat from $25. I read the US pricing page but tested on the owner's Team plan and did not open a checkout, so pricing counts as 3.
Prices read on 12 November 2026. Paying yearly lowers Pro to $17 a month and a Team seat to $20.
Claude plans and prices
Live data· last check 12 November 2026| Plan | Price a month | |
|---|---|---|
| FreeLimited usage every 5 hours and every week; web search, file creation and code | $0 | Limited usage every 5 hours and every week; web search, file creation and code |
| Claude ProMore usage than Free, Claude Code included | $20 | More usage than Free, Claude Code included |
| Claude Max5x the Pro usage, Claude Code included, $100 a month in API credits | $100 | 5x the Pro usage, Claude Code included, $100 a month in API credits |
| Claude TeamStandard seat for teams of 2 to 150, more usage than Pro | $25 | Standard seat for teams of 2 to 150, more usage than Pro |
- Prices exclude applicable tax
- Enterprise is $20 per seat per month plus API-rate usage, billed annually; contact sales (not listed here)
- Paid plans can turn on usage credits at standard API rates when limits are hit
On the Team plan the message box said "Now using credits" for my first three messages, because they ran just before a plan reset. After the reset the notice did not come back.
Claude Accuracy and Invented Facts
- Answer Accuracy & Hallucination
- 5/5Excellent
Claude answered all 30 fact questions correctly and invented nothing. It rejected every question with a false premise, such as a novel Dickens never wrote and a telescope Edison never built.
Every answer came from Opus 5.5 at Medium effort on a Team seat, so that is the model you can expect these results from.
Claude Reasoning and Maths
- Reasoning & Maths
- 5/5Excellent
Claude solved all 15 reasoning and maths problems on its first answer.
Fifteen problems is a small set, so if you need more proof you may want a harder set to separate the top chatbots.
Claude Coding
- Coding
- 5/5Excellent
Claude wrote code that passed all ten of my checks the first time, with no fixes from me, so you can paste it and run it. That includes a task to write its own tests, which caught every broken version I fed it.
I kept only the code, so I cannot tell you how well it explains its work to you, or whether it flags its own assumptions.
Claude Following Instructions
- Following Instructions & Writing
- 5/5Excellent
Claude met 49 of 50 constraints. A 60-word summary came back at exactly 60 words, and a sentence that must have 10 words had exactly 10.
It named the contradictory brief as impossible and gave one text of each length. Its one miss was an email that began with a subject line before the greeting I asked for.
Claude Memory and Long Documents
- Context & Long-Document Memory
- 5/5Excellent
Claude read a 14,786-word file whole and found all ten planted facts. You can upload a long report and ask it questions about it.
In a 25-turn conversation it kept every number and name straight, and it ended every reply with the word I asked for, to the last turn. It also pointed out that the date I gave had already passed, and asked whether I meant next year.
Claude Images, Files and Voice
- Images, Files & Voice
- 4/5Good
Claude passed eight of nine image, file and audio points I could run. It read a photo, a chart, handwriting, a PDF and a spreadsheet, and it said so when a PDF did not name a person.
The audio clip failed: Claude said it cannot listen to audio and could not transcribe the MP3. The two image requests came back as vector drawings, so you get illustrations and not photos, and I could not test voice chat, because it needs a microphone.
Claude Safety and Refusals
- Safety & Refusals
- 3/5Average
The first safety prompt was paused: the page showed that Opus 5.5's safeguards flagged the message, and I sent no more.
The notice says this sometimes happens with safe, normal conversations. It was a pause on one message and not a warning on the account.
Claude Response Speed
- Response Speed
- 5/5Excellent
Claude showed its first words in under one second and finished in about two, at the quick end of the scale. The slowest answer took about eight seconds.
No answer failed or timed out, so nothing left you waiting. The vector drawings took up to 38 seconds. I timed with a script, so treat the figures as close, not exact.
Final verdict: is Claude worth it?
Claude is worth using if you want an accurate, fast chatbot for facts, maths, code and long documents. It got every fact, reasoning problem and code task right and kept every rule I set.
It is a weaker fit if you need audio transcription or photo-style images in the chat. My account had no web search, and the safety test paused at its first prompt, so I cannot say how it handles live facts or harmful requests.
The nine tests that count average 4.4 out of 5, with safety and pricing counted as 3 because I did not run them. I added nothing, because a Team plan and a missing web search tool limit what I can say, so I rate it 4.4.
See the Claude test results and how I rate.
Best for: People who want a very fast, accurate chatbot for facts, maths, code and long documents.
Video review · 3:01
Watch me test Claude
The whole test in about three minutes: facts, reasoning, code, writing briefs, a long document, images, files and speed. Web search was off on my account. Captions and the full transcript are one click away, and the video has a page of its own.
Transcript
Hi, I'm Miles. I tested Claude with the same fixed battery I give every chatbot, on the owner's Team plan. Here's the short version. It got every fact, every reasoning problem and every code task right, and it kept every rule I set. The catch is my account, which had no web search, and the safety test stopped at its first prompt. My rating is four point four out of five.
So what is Claude? It's Anthropic's chatbot, in a web app. You ask questions, upload files and get written answers. It doesn't make photo-style images in chat. My account showed one model, Opus five point five, and it had no web search tool.
Every chatbot gets the same questions, and I keep the first answer to each. Seven tests are graded. Safety and pricing count as three, and I left the live web test out, because my account had no web search. Every test ran in a fresh incognito chat.
On facts, Claude got all thirty right and invented nothing. It rejected every question built on a false premise. It also solved all fifteen reasoning problems, and all ten code tasks passed first time, with no fixes from me.
On writing briefs it met forty-nine of fifty constraints. The sixty word summary came back at exactly sixty words. It named the contradictory brief as impossible and gave one text of each length. Its one miss was a subject line before the greeting I asked for.
I uploaded a document of nearly fifteen thousand words with ten facts hidden in it, and Claude found all ten. In a long conversation it kept every number straight and ended every reply with the word I asked for, right to the last turn.
It read a photo, a chart, handwriting, a PDF and a spreadsheet correctly. The audio clip failed, because Claude said it cannot listen to audio. It drew illustrations instead of making photo-style images, and I could not test voice chat, because it needs a microphone.
The first safety prompt was paused. The page said Opus five point five's safeguards flagged the message, and I sent no more. So I counted safety as three, and I can't tell you more than that.
Claude showed its first words in under one second, and finished in about two. The slowest answer took about eight seconds. That is the quick end of my scale.
There is a free plan and paid plans, Pro, Max and Team. I read the pricing page, but I tested on the owner's Team plan and did not open a checkout, so I counted pricing as three. The prices on the page come from my price tracker.
So, is it worth using? Yes, if you want an accurate, fast chatbot for facts, maths, code and long documents. Audio and photo-style images are the gaps, and I could not grade live web or safety. The nine tests average four point four, and I add nothing. My rating is four point four out of five.
Chapters · 13
Frequently asked questions
Is Claude free?
Yes. The US pricing page lists a free plan at $0 and paid plans from $20 a month. I tested on a Team plan, so I did not measure the free plan's limits.
Is Claude accurate?
Yes. On what I could count it got 30 of 30 fact questions and 15 of 15 reasoning problems right. My account had no web search, so I did not grade live answers.
Which Claude model did I test?
Opus 5.5. The message box showed it at Medium effort on the Team plan, and I never changed it.
Can Claude make images or listen to audio?
No. In my test it drew vector illustrations and said it cannot make photo-style images, and it could not transcribe an MP3.
What happens when Claude hits a limit?
It uses credits. For my first three messages the box said "Now using credits" until the plan limit reset, and the notice did not come back.
Miles FosterOwner and tester, RankDudeMiles Foster is a pen name: the owner of RankDude, a private person, runs every test himself.
Disclosure. I tested on a Claude plan RankDude paid for at the public price. Claude didn’t see this review first and had no say in my rating. How RankDude makes money is on the disclosure page.