# FinGPT

Financial LLM adapters, instruction datasets and training notebooks, all MIT.

*https://stockmarketstack.com/tools/fingpt · Financial MCP Servers & AI Agents*

## Facts

### At a glance

| Field | Value |
| --- | --- |
| Vendor | AI4Finance Foundation |
| Category | Financial MCP Servers & AI Agents |
| Job | infrastructure |
| Website | https://fingpt.io |
| Pricing model | open-source |
| Free tier | true |
| Open source | true |
| Licence | MIT |
| Self-hosted | true |
| Tested hands-on | false |
| Last updated | 2026-09-13 |

### Coverage

| Field | Value |
| --- | --- |
| Asset classes | stocks |
| Markets | us, asia |
| Works outside the US | true |
| Data latency | none |
| Platforms | library |
| AI features | recommendations |

### Interfaces

| Field | Value |
| --- | --- |
| API | false |
| Webhooks | false |
| Scripting | none |
| Python | true |
| Spreadsheet add-in | false |
| MCP server | false |
| Export | none |

### Capabilities

Yes: none

No: charting, screening, scanning, backtesting, automation, live_trading, paper_trading, portfolio_tracking, broker_import, tax_reporting, alerts, news, options_analysis

*Verified: pricing 2026-09-13; capabilities 2026-09-13; coverage 2026-09-13.*

## What it is

FinGPT is the AI4Finance Foundation's open financial LLM project, and what you download is
weights and data rather than an application: eleven LoRA adapters on Hugging Face — fine-tunes of
Llama-2, Llama-3, ChatGLM2, Falcon, MPT, Bloom, Qwen and InternLM for sentiment, named-entity
recognition, relation extraction and headline classification — fourteen instruction datasets, and
the notebooks that trained and benchmarked them.

The repository is alive; the modelling is not. MIT per its LICENSE file, 21.2k stars, not
archived, last commit to master on 12 September 2026, one tagged release (v1.0.0, 8 April 2026) —
but the newest adapter was last touched in October 2024 and the rest are from 2023 and the first
half of 2024. Over the past 30 days the instruction datasets pulled thousands of downloads, 2,015
for fingpt-sentiment-train alone, while eight of the eleven adapters pulled none. People take the
data and train their own.

## Pricing

The repository and the weights are MIT, commercial use included. The project site advertises "API
& Service Plans" on top — Free at $0 capped at 100 sentiment requests a day, Premium at $19.90 a
month for 10,000, Enterprise on request — with nothing behind them: no API host resolves, the
documentation page's "API Reference" leads back to the README, the advertised demo path returns
404, and "Start Premium" is a jump link to a contact block whose only action is an email address.
Budget for GPU time, not for a plan.

## Data & coverage

FinGPT ships no market data. Its datasets are labelled financial text — sentiment over news and
tweets, headline classification, NER, relation extraction, FiQA Q&A and a Chinese
multiple-choice set — covering US news and Chinese A-share material. The most application-shaped
piece, FinGPT-Forecaster, is a Llama-2-7b LoRA trained on Dow 30 news to 2023 that emits a
next-week price-movement call from Finnhub and yfinance data you fetch with your own keys — its own
README calls it a "junior robo-advisor", which is why this card sets `ai_features: recommendations`.

## Integrations

Python, and a 2023 Python: the Forecaster pins torch 2.0.1, transformers 4.32.0, peft 0.5.0 and
gradio 3.50.2, and the setup guide warns off 3.12. Llama-2 base weights are gated, so you also
need a Hugging Face token. The README advertises a no-GPU path — set `FINGPT_LLM_PROVIDER=openai`
and import `fingpt.Forecaster` — but no such module exists in the repository, and the only code
that reads that variable belongs to an unrelated sub-project committed alongside FinGPT.

## Limitations

- The published weights are two years old, on base models that have been superseded. What is
  still worth having is the datasets and the training recipe, not the adapters.
- The hosted demo does not run: the Forecaster Space sits in a runtime error — importing `peft`
  pulls in `bitsandbytes`, which cannot import `driver` from `triton.runtime` — and the
  Chinese-language Space is paused.
- No packaging worth the name — the PyPI entry is a stub at 0.0.1 from October 2023, with no
  console entry point — so you vendor the repository or copy the notebook.
- Fine-tuning needs an NVIDIA card with 12GB+ of VRAM; the headline v3.3 sentiment model took
  17.25 hours on one RTX 3090.
- Nothing here is inference infrastructure: no server, no API, no evaluation harness you would
  put in front of production traffic.

## Alternatives

FinBERT is smaller and simpler if sentiment is all you want. [FinRobot](https://stockmarketstack.com/tools/finrobot), from the same foundation,
is the thing that runs rather than the thing you fine-tune. And a current general-purpose model
with a decent prompt will beat these 2023 adapters on most of the tasks they were built for.

## FAQ

### Is FinGPT software I can install, or a research project?

Mostly the latter. What you get is eleven LoRA adapters and fourteen instruction datasets on Hugging Face, plus the notebooks that trained and benchmarked them. There is no usable package — the PyPI entry is a stub last published at version 0.0.1 in October 2023 — so the install is a git clone and an editable pip install, and the thing you run afterwards is a notebook.

### What hardware does running a FinGPT model need?

The adapters are LoRA weights on 7B to 20B base models, so you load the base model yourself. The project's own setup guide asks for 8-16GB of RAM and 20GB of disk for inference, and an NVIDIA card with 12GB+ of VRAM and CUDA 11.8 for fine-tuning. Its headline result, the v3.3 sentiment model, took 17.25 hours on one RTX 3090.

### Are the FinGPT models still being updated?

No. The newest adapter on Hugging Face, fingpt-mt_llama3-8b_lora, was last touched in October 2024; the rest date from 2023 and the first half of 2024. The repository itself still takes commits — the last was 12 September 2026 — but they are bug fixes, documentation and new side-projects, not new weights.

### Is FinGPT the same thing as FinRobot?

No, and they are licensed differently. FinGPT is an MIT repository of fine-tuned financial LLMs and datasets. FinRobot is the same foundation's Apache-2.0 multi-agent platform, with a desktop build and a hosted version whose source is not open. FinGPT is what you fine-tune; FinRobot is what you run.

## Also from AI4Finance Foundation

- [FinRobot](https://stockmarketstack.com/tools/finrobot.md)

## Also worth comparing

- [AI Hedge Fund](https://stockmarketstack.com/tools/ai-hedge-fund.md) — LLM investor personas score a ticker, deterministic code sizes the book, nothing trades.
- [MaverickMCP](https://stockmarketstack.com/tools/maverick-mcp.md) — Thirty-seven stock-analysis tools for any MCP client, on Yahoo data and no key.
- [SEC EDGAR MCP](https://stockmarketstack.com/tools/sec-edgar-mcp.md) — SEC filings and XBRL financials as MCP tools, under AGPL or a paid commercial licence.
- [TradingAgents](https://stockmarketstack.com/tools/tradingagents.md) — Analyst, bull, bear, risk and portfolio-manager agents argue over one ticker.
- [Vibe-Trading](https://stockmarketstack.com/tools/vibe-trading.md) — A local finance research agent with 74 MCP tools, nine backtest engines and 14 brokers.
- [Equibles](https://stockmarketstack.com/tools/equibles.md) — Scrapes SEC, FINRA, FRED and CFTC data into your own Postgres and serves it over MCP.