Compare / On-chain analytics with execution / Verified 3 September 2026
Fere AI vs Nansen
A source-linked comparison of Nansen and Fere AI: the deepest wallet-label dataset in crypto against an agent that keeps running after you close the tab.
Fere AI vs Nansen
Nansen is an on-chain analytics platform built on more than 500 million labelled addresses, with an on-chain AI analyst, a non-custodial wallet powered by Privy, and spot and perps trading from the agent or the terminal. Fere AI compiles a sentence into readable rules and runs them on a schedule under hard caps. Nansen added execution, so these two now overlap directly — but they are organised around different questions. Nansen is built around research you drive: follow a wallet, see what Smart Money bought, then trade it. Fere is built around a strategy that keeps running after you close the tab. For a lot of traders the answer is both, and this page will not pretend otherwise.
All claims about Nansen below come from Nansen's own site and documentation, verified 3 September 2026. Sources are linked at the bottom of this page.
Why traders choose Fere AI
- It works when you are not there. A Fere strategy runs on a schedule, reports each run in a sentence, and stops at a cap you set. It does not need you to notice an alert.
- A staged path to live. Monitor, dry run, paper, capped live. You promote each step, so a new rule never starts with real money.
- One cap per strategy. Each strategy holds its own wallet and its own limit rather than sharing a single account balance.
- Drivable from your own assistant. The MCP server lets Claude or ChatGPT research and trade through Fere directly.
At a glance
| Capability | Nansen | Fere AI |
|---|---|---|
| What it is | An on-chain analytics platform built on 500M+ labelled addresses, with an AI analyst and spot and perps trading | An agent platform that turns a sentence into readable rules and runs them on a schedule under hard caps |
| Wallet-label intelligence | 500M+ labelled addresses with Smart Money, whale, fund and public-figure tracking. This is the core asset and the reason to buy | None. Fere does not publish a labelled-address dataset and does not compete on this |
| How you decide what to trade | You drive it — follow wallets, read the analyst, then act on what you find | One sentence. Fere asks at most two clarifying questions, then compiles a spec you can read — numbered rules in plain English. |
| What happens before it goes live | You place the trade when you decide to; no staged rollout is documented | A promotion ladder. Monitors risk no funds; trading strategies dry-run once, then paper-trade, then go live under hard caps. You promote them, never the other way. |
| Who holds the keys | Non-custodial wallet powered by Privy, with deposit and withdraw from the platform | Non-custodial. Every strategy gets its own capped wallet, keys never leave an AWS Nitro TEE, and you can export to MetaMask or Phantom. |
| What it executes | Spot and perps, from the agent or the terminal | On-chain across Ethereum, Base, Arbitrum, Polygon, HyperEVM, Solana and BSC, with perps on Hyperliquid. Routed through Jupiter, 0x and LI.FI — best quote wins. |
| Runs unattended | Their site leads with alerts and agent-assisted trading rather than scheduled autonomous strategies; no scheduled-strategy product documented as of 3 September 2026 | Yes. Strategies run on a schedule with hard position caps and a kill switch on every screen, and each run is explained back to you in a sentence. |
| Inside ChatGPT and Claude | An API is offered; no MCP server documented as of 3 September 2026 | Yes. An MCP server at api.fereai.xyz/mcp — add the connector to Claude, ChatGPT, Cursor or any MCP client, authorise once with OAuth, and the wallet is provisioned automatically. |
| Best for | Traders whose edge is wallet-level research — tracking Smart Money, whales, funds and public figures across chains | Traders who can describe a strategy in a sentence and want it running unattended, with a staged path to live and a cap on what it can lose. |
| Scope | The loop stops at a decision you have to make. Nothing described keeps executing a rule for weeks without you | Strategy automation end to end: authoring, dry run, paper, capped live execution and scheduling, on-chain across seven chains. |
Where they differ
One loop ends at a decision, the other at a cap
Nansen's loop is discovery to execution: see what Smart Money is buying, trade it, track the position. That loop is fast and it is genuinely theirs — no one else has the label graph to make step one work. But it ends with a decision you make, every time. Fere's loop ends somewhere else: a rule you wrote once keeps running on a schedule, reports each run back in a sentence, and stops itself at a cap. If your edge is spotting things, Nansen extends your reach. If your edge is a rule you already trust, Fere is what runs it while you sleep.
Two different sources of edge
The two products draw their edge from different places. Nansen's comes from address labelling at scale — if the question you need answered is which fund just bought a token, that is the dataset built for it. Fere's comes from market data, technical indicators and web and social sources, compiled into a strategy that then runs. They are answering different questions, and a desk that wants both signals is not making a mistake by running both.
Where the strategies come from
Fere reads traders, desks and forums continuously and compiles what it finds into strategies you can run, fork or blend, each credited to its source and carrying a thirty-day backtest. It is a different answer to the same underlying question Nansen answers with labels — what should I be doing right now — arrived at from published reasoning rather than from wallet flow. The two are complementary far more than they are substitutes, and running Nansen for discovery while Fere runs the rules is a perfectly coherent setup.
Nansen strengths
- The label graph. 500M+ labelled addresses with Smart Money, whale, fund and public-figure tracking. Nothing else in crypto comes close, and it is the whole reason the product works.
- Discovery to execution in one place. Seeing a flow and trading it without leaving the platform is a real workflow advantage.
- Perps as well as spot. Both are tradeable from the agent or the terminal.
- A large published research corpus. Their explainer library is substantial and well ranked, which also means it is easy to evaluate them before paying.
- Non-custodial wallet infrastructure. Powered by Privy, with deposit and withdraw handled inside the platform.
- An API. If you want to build on their data rather than read it, that path exists.
Fere AI strengths
- Strategies that run unattended. Scheduled execution with hard caps and a kill switch, each run explained in a sentence.
- The promotion ladder. Monitor, dry run, paper, capped live — a rule never starts with real money.
- Sourced strategy library. Run, fork or blend strategies compiled from traders and desks, each credited and backtested over thirty days.
- Per-strategy capped wallets. A bad rule can only spend what that rule was given.
- MCP access. Claude, ChatGPT or Cursor can research and trade through Fere without opening the app.
What each one is built for
| What Nansen is built for | What Fere AI is built for |
|---|---|
| Your edge comes from knowing which wallets are moving. | You want a rule to keep running when you are not watching. |
| You want to see a flow and trade it in the same session. | You want a staged path from idea to live money. |
| You want deep on-chain analytics as the primary product, with execution attached. | You want a cap scoped to each strategy rather than an account balance. |
| You are researching across chains at address level rather than running rules. | You want to drive research and execution from Claude or ChatGPT. |
Fere AI pricing
Lite is $9 per month (1,100 credits), Pro is $29 per month (4,200 credits), and Whale is $99 per month (16,000 credits). Annual billing is two months free compared with paying monthly. Every tier includes the full product; credits are the usage limit. We do not republish Nansen's prices here — check them at nansen.ai.
The short version
Nansen is built around wallet intelligence; Fere is built around strategies that keep running. Nansen's label graph answers what other wallets are doing. Fere answers what your own rule should do next, on a schedule, on a ladder, under a cap. They sit at different points in the same workflow, and plenty of desks will sensibly run both.
How this comparison was written
Every claim on this page about the other product comes from their own site or documentation, checked on the date shown, with the sources linked below. We describe what each product is built for rather than guessing at its roadmap, and we do not republish competitor prices — those change without notice, so check them at the source.
Sources, verified 3 September 2026: nansen.ai.
Other comparisons
- Fere AI vs Minara — AI trading agent
- Fere AI vs Bankr — Social trading agent
- Fere AI vs PayBox — Agent payment vault
- Fere AI vs FOMO — Social mobile trading app
- Fere AI vs Surf — Market intelligence platform
- Fere AI vs TrueNorth — AI research and trade planning
- All comparisons
Still deciding? Start with a sentence.
Describe what you want to watch or trade. Fere compiles it into rules you can read, dry-runs it once so you see the exact order it would have placed, and only goes live when you promote it.