Your Mac. Your superpowers.
For investors, researchers and traders: research companies, test trading ideas on paper and watch the markets, with AI running on your own Mac. The decision stays yours.
X-Ray Vision
See through any market move.
You see the price. Understanding what sits behind it takes work.
Look through a move into the company’s figures, filings, earnings calls and news, with every piece of evidence open for you to inspect.
On your Mac: the AI reads the report and marks the sentence that explains the move.
Flash
From question to answer in a flash.
Gathering information eats the time you could spend thinking.
Ask once. London5 turns your question into connected research, calculations and charts, so you reach your decision sooner, with time left to think.
On your Mac: four documents read, two calculations and a chart, for one question.
Pattern Vision
See the links others miss.
Important relationships disappear across separate tools and timeframes.
Compare companies, sectors and periods side by side. London5 shows the connections worth investigating, with the evidence behind each one.
On your Mac: thousands of products and stories compared, five years back.
Time Travel
Rewind the market. Test your idea.
You can’t live through years of market history before you test an idea.
Replay years of past prices and run your strategy against simply holding. See the returns, the fees and the worst drop, on paper money.2
On your Mac: two years of daily prices replayed in under a minute.
Market Radar
Keep every market on your radar.
You can’t watch every market all day.
Set watches, follow your symbols, and London5 brings the changes that matter to your attention, and lets the rest go by.
On your Mac: your watches keep running while your Mac stays on.
Parallel Minds
Explore many paths at once.
Investigating several companies, or competing explanations, stretches your attention.
Look into several companies, questions and scenarios at the same time. London5 runs them side by side and brings the findings together for your review.
On your Mac: five research runs at once, nearly a hundred documents.
The power source
A dedicated AI, on your own chip.
London5 runs open-source AI models on the computer you already own. We test them on real machines and recommend the ones that fit yours.
Apple M5 family
macOS · available now
Speeds measured on an M5 Pro with 48 GB of unified memory
AMD Ryzen AI Max+ 395
Windows and Linux · coming soon
Up to 128 GB of unified memory, room for the largest models
Running on this chipQwen3.8 27B · 3-bit · 11 GB · open source, Apache 2.0
- Speed on the M5 Pro
- 17.7 tokens/s
- Questions today
- 16
- AI cost
- €0.00
Open-source models we testspeed measured on the M5 Pro, 8 Sep 2026
| Model | Needs | Measured |
|---|---|---|
| Qwen3.5 4B | 8 GB | best tool use of 13 models: 97.5% |
| Qwen3.8 27B · 3-bit | 24 GB | 17.7 tokens/s |
| Nemotron 3.5 Lightning 30B | 32 GB | 68 tokens/s |
| Qwen3 32B | 48 GB | 11.6 tokens/s |
Chosen on measured tool use, not size. Models above 48 GB are measured next, on 128 GB.
Stored on this computernot on our servers
- Conversations
- Research
- Simulations
- Watches
With a local model the AI runs here too, so your questions never leave the machine. Prices and news still come in from the internet.
Try LONDON5 free for 60 days1
Download London5 and use all of it: research with sources, the simulator, watches and your choice of AI model.
Try it free for macOSCreate your London5 account with your email, and we’ll email you a code to download the app. Requires macOS 14 Sonoma or later; Apple silicon recommended, Intel Macs supported.
macOSAvailable now
Apple silicon and Intel. 8 GB of memory minimum; 16 GB to run a local AI model.
WindowsComing soon
Windows 11 or later. An NVIDIA GPU is recommended for a local AI model.
LinuxComing soon
Ubuntu 22.04 or equivalent. An AMD or NVIDIA GPU is recommended for a local AI model.
- Cloud AI and data providers you choose may charge you separately. London5 charges no per-trade or performance fees.
- Simulations run on paper capital. A result on past prices doesn't tell you what comes next, and it isn't a recommendation.