Now Open

Early in your career and hungry to build something real? We're looking for driven interns — no experience required, just curiosity and commitment.Internships open — curiosity over credentials.

/Fincept
EnterpriseProductsPricingCompareOpen SourceQuantcept
Resources
Terminal ManualBlogDocumentationFAQGitHub
Company
About UsTeamCareersContact
Schedule DemoLoginGet the Terminal

The Open-Source Fincept Terminal

A native C++20 · Qt6 · embedded Python 3.11 desktop terminal, open on GitHub under AGPL-3.0 with 30k stars and 4.2k forks. Free for personal, learning and academic use. This page is what it includes, how it is licensed, and where Enterprise picks up.

Download Desktop AppLaunch Web TerminalView Source
30k
GitHub Stars
4.2k
Forks
AGPL-3.0
Open Licence
100+
Data Connectors

Three Ways to Run the Open Edition

The same core data workflow, whichever way you reach it. None of these needs a card — but read the licence above before running any of them at work.

Desktop

FinceptTerminal Desktop

A native C++20 / Qt6 binary with embedded Python analytics. Signed installers for Windows x64 (.exe), Apple Silicon macOS (.dmg) and Linux (.deb, .rpm or portable .run) — or build it yourself from source with CMake, or run the Dockerfile.

Download Latest Release →
Browser

Hosted Web Terminal

The same terminal with nothing to install — markets, watchlists, portfolio, research, news and economics in any modern browser. A free account is all it takes.

Launch Web Terminal →
API

Financial Data API

The REST API behind the terminal — market data, fundamentals, news, economics, symbol search and QuantLib pricing. Free tier included; credit packs below for more volume.

Read the API Docs →

What the Open Edition Includes

Everything below is in the free build, on every platform, with no usage tier attached.

37 AI agents across trading, economics and geopolitics
100+ data connectors — FRED, IMF, World Bank, DBnomics, AkShare, Polygon, Kraken
QuantLib suite — 18 modules for pricing, risk, stochastic and fixed income
Multi-asset analytics — DCF, portfolio optimisation, VaR, Sharpe, derivatives
Real-time crypto and equity trading with 16 broker integrations
Visual node editor for automation pipelines, plus MCP tool integration
AI Quant Lab — ML models, factor discovery and reinforcement-learning research
Maritime tracking, geopolitical analysis and satellite data
Bring your own LLM — OpenAI, Anthropic, Gemini, Groq, DeepSeek or local Ollama

Open Source, Not Permissive

The open edition is published under AGPL-3.0. That is a strong copyleft licence with real obligations attached — the LICENSE file in the repository is the authority, and worth reading before you build on it.

IndividualsFree

Personal use, learning and academic research. No charge and no card — download the build and run it. Keep in mind the AGPL network clause: distribute a modified version, or run one as a service others reach, and your changes have to be shared under the same licence.

Teams & FirmsFrom $99/mo

For a desk, take Fincept Terminal Enterprise instead. A clean commercial subscription with no copyleft obligations to manage, plus private data, multi-agent research, the dataroom, SSO and audit logs — and it is the edition we actively develop.

Before You Choose the Free Edition

It is a genuinely capable terminal and we are not going to pretend otherwise. But if you are doing this for a living, six things are worth knowing before you commit a workflow to it.

01

It is not actually free

Open source

You supply your own API keys for the commercial data providers and your own LLM key for every AI feature. That inference is billed to you directly, per token, with no ceiling — a busy research month is real money to OpenAI or Anthropic, on top of whatever the data costs.

Enterprise

Credits are included in the seat — 400 to 5,000 a month — and the proprietary data is licensed by us. One predictable line item instead of three variable ones, and the terminal tells you what each action costs before you run it.

02

Development happens here

Open source

The public build now ships roughly one update a month. It is maintained, not abandoned — but it is not where the product is going.

Enterprise

The private edition is where the work happens, and fixes and features arrive as they land rather than waiting for the next public cut.

03

Your issues are triaged first

Open source

Community reports go into the public queue and are addressed on a best-effort basis. There is no response commitment, and a fix reaches you on the next monthly build.

Enterprise

Priority support with an SLA. Your report is triaged ahead of the public queue, and the fix ships to you when it is ready — not on the public release cadence.

04

The data underneath is different

Open source

Public sources — good enough to see a market, rate-limited and best-effort, and capped by whatever the upstream provider gives a free key.

Enterprise

Private and proprietary datasets with deeper history and faster refresh, licensed for Enterprise accounts. Price history runs to five years or the full archive rather than one.

05

The ceilings are where the work stops

Open source

Practical limits are set by the free tiers of whatever you connect: a handful of symbols, shallow history, and quotas you do not control.

Enterprise

Up to unlimited portfolios, watchlists and strategies, 500 streaming symbols, 53 agents, 6 team modes and unlimited MCP servers. Published limits, enforced by us, that you can plan a workflow around.

06

The licence is a compliance question

Open source

AGPL-3.0 is strong copyleft. Modify it and run it as a service others reach, and you are obliged to publish your changes under the same licence — which for a fund means your internal tooling.

Enterprise

A clean commercial subscription with no copyleft obligations to reason about. For most legal teams this is the row that ends the discussion.

Add up your data keys and a month of your own LLM spend before deciding the free edition is the cheaper option. For most people doing this professionally, it is not.

See What a Seat CostsExplore Enterprise

Open Source vs Enterprise

The honest split. The open edition stays free and maintained; the Enterprise edition is where private data, the agent and quant modules, and the compliance controls live.

Feature
Open Source
Enterprise
Licence
AGPL-3.0 + paid commercial
Commercial subscription
Account required
Free Fincept account
Enterprise account
Where it runs
Desktop app + browser
Desktop app
Market data
Public sources
Private + proprietary
Usage limits
Standard
Greatly expanded
AI research
Standard agents
Multi-agent, planning
Live broker routing
Live algo deployment
Point-in-time data
Priority support & SLA
Update cadence
Monthly
Active development
Support
Community
Dedicated
Price
Free for individuals
$99+ / user / mo

Need the private data, the research agents or SSO? See Fincept Terminal Enterprise — from $99 per user per month.

Credit Packs for the Data API

For programmatic access to the Fincept REST API. Buy once, use at your own pace — these are one-time purchases, not a subscription, and they are separate from Enterprise seats.

Loading credit packs...

Looking for terminal seats instead? See Enterprise pricing →

Built in the Open

Issues, pull requests and forks are welcome. The codebase is public so compliance teams can audit what the terminal does with data, and quant teams can extend it.

Contribute on GitHubReport an IssueRead the Manual

Open-Source Questions

AGPL-3.0 — the GNU Affero General Public License v3.0. It is not MIT and it is not permissive: it is strong copyleft, which means modifications have to be shared under the same licence if you distribute the software or run it as a network service. Read the LICENSE file in the repository before you build on it.

For a team deployment we point you at Fincept Terminal Enterprise rather than the open build — it starts at $99 per user per month, comes with a straightforward commercial subscription with no copyleft obligations, and includes the licensed data, the AI credits and priority support with an SLA. We no longer sell a standalone commercial licence for the open edition.

AGPL-3.0 is strong copyleft, not a permissive licence. Running it privately for yourself carries no obligation. But if you distribute a modified version, or run one as a service other people reach over a network, you have to make your source changes available under the same licence. That network clause is the part teams most often miss. The LICENSE file in the repository is the authority — read it before deploying.

Yes — individuals using it for their own account, for learning, or for academic research. That is who the open edition is for, and there is no charge and no card. Teaching or campus deployments are worth a conversation; get in touch and we will work out what fits.

Yes, but at a reduced cadence, and we would rather say so than let you find out. The public terminal now ships roughly one update per month rather than continuous daily development, because the team is focused on the subscription-funded private edition. Everything already released stays released, issues and pull requests are still reviewed, and the repository is not being archived.

The current generation is a native desktop application: C++20 with Qt6 for the interface and rendering, and an embedded Python 3.11 runtime for the analytics layer. It ships as a single native binary rather than a Python package, which is why you install it from a platform installer instead of pip.

Four routes. A signed platform installer is the easy one — .exe for Windows x64, .dmg for Apple Silicon macOS, and .deb, .rpm or a portable .run for Linux. There is also a setup script in the repo, a Dockerfile if you would rather containerise it, and a full CMake build from source with pinned toolchain versions if you want to compile it yourself.

For some sources, yes. The terminal ships connectors for a large number of providers — public ones like FRED, the IMF, the World Bank, DBnomics, AkShare and government statistical APIs work out of the box, while commercial feeds and broker connections need your own credentials for that provider. A free Fincept account issues a Fincept API key for the endpoints we serve directly.

Yes, and this is one of the better reasons to run the open build. It supports multiple providers — OpenAI, Anthropic, Gemini, Groq, DeepSeek, OpenRouter and others — as well as local models through Ollama, so you can keep inference entirely on your own hardware if that is what your policy requires.

No. Every credit pack is a one-time purchase — you pay once and receive a fixed number of credits, with no recurring charge and no auto-renewal. Each pack has a validity window shown on its card, counted from the date of purchase.

Major credit and debit cards, net banking and UPI, processed by Cashfree — a PCI-DSS compliant gateway. Card details never touch our servers.

Yes — issues and pull requests are welcome, and the repo carries separate contributing guides for the C++ and Python sides. Contributions are accepted under the same AGPL-3.0 terms as the rest of the codebase.

Yes, but it is a new account rather than an upgrade. Enterprise runs on a separate backend with its own login, so you create an Enterprise account and subscribe there. Your Fincept account and any API credits on it are unaffected and keep working.