whatmodel.sh / Falcon 40B Instruct

Falcon 40B Instruct

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Falcon 40B Instruct is an open-weights language model from TII, released May 2023. Its published size is 41.8B parameters. Licence: apache-2.0.

Catalogue identity notes

tiiuae/falcon-40b-instruct @ecb78d9 (created 2023-05-25, licence apache-2.0). TII's 2023 instruct fine-tune of Falcon-40B (catalogued base falcon-40b). Independent rows: AlpacaEval 1.0 and 2.0 and LMSYS MT-Bench.

Model information sources

Release date ↗ · Weights licence ↗

No safetensors metadata (pytorch .bin weights); ~41.8B per the ollama tag page ('parameters 41.8B').

28.2±2.1#656 of 719 LLMsfrom 4 benchmark results across 3 sources

$ how it compares

The LLMs ranked either side of Falcon 40B Instruct, out of 828 with a score, then more from TII.

#modelscorewhereRAMreleased
753NexusRaven V2 13B Nexusflow29±3local~9.3GB2023-12
754OLMo 7B Allen Institute for AI28±1local~30GB2024-02
755Falcon 40B Instruct TII28±2local~27GB2023-05
756GPT-NeoX-20B EleutherAI28±1local~14GB2022-02
757LLaMA 13B Meta28±1comparison only2023-02
more from TII
233Falcon-H1R 7B TII60±1local~6GB2026-01
332Falcon-H1 34B Instruct TII54±3local~23GB2025-05
494Falcon-H1 7B Instruct TII46±3local~6GB2025-05

see all LLMs · find the ones that fit your machine

$ what it's good at

chat & writing▁▁▃▃▅▄▅█▆▆▇▃ #384 of 393 · 95% of its score · 3 tests

Beats Phi-2 (2.7B) here, though they rank higher overall. Trails Gemma 3 270M IT, SmolLM2 360M Instruct, Granite 4.0 350M here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
AlpacaEval 2 length-controlled5.6% ✎▁▂▃▃▆▆█▆▄▄▁▁ beats 10/122medium · avg 39%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 37%too hard for it
MT-Bench (GPT-4 single-answer grading, LMSYS runs)5.1700 MT-Bench score (1-10) ✎▆▂▂▆▆▄██▂▄▄▆ beats 9/28avg 8.819 MT-Bench score (1-10)▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 34%sharp · noisy
AlpacaEval 1.0 win rate (vs text-davinci-003, GPT-4 judge)45.7% ✎▂▂·▆·▇▃▆█▆▄▂ beats 5/45very easy · avg 96%▇▇▇▇▇▇▇▇▇▇ 24%sharp · at its level · noisy
other tests1 not in a task · 3.2% of its score
testresultvs the others in its rundifficultycountswhy
TruthfulQA MC252.5% ✎▅▃█▆▁▃▅···▁▁ beats 24/31medium · avg 64%▇ 3.2%blunt · at its level

$ where to run it

llama.cpp · recommended

version

  1. Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
    macOS / Linux
    curl -LsSf https://llama.app/install.sh | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run Falcon 40B Instruct · needs ~26 GB · from huggingface.co/mradermacher
    chat in the terminal
    llama cli -hf mradermacher/falcon-40b-instruct-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf mradermacher/falcon-40b-instruct-GGUF:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run Falcon 40B Instruct · needs ~26 GB · from huggingface.co/mradermacher

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/mradermacher/falcon-40b-instruct-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Falcon 40B Instruct · needs ~24 GB · from ollama.com
    chat in the terminal
    ollama run falcon:40b

Official weights and documentation: huggingface.co/tiiuae ↗

$ how the score was worked out · AlpacaEval 1.0 and 2.0 leader… 61%, LMSYS Chatbot Arena leaderboa… 34%, Hallucinations Leaderboard re… 3%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
AlpacaEval 1.0 and 2.0 leaderboards (tatsu-lab/alpaca_eval repository)2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 61%Apache-2.0
LMSYS Chatbot Arena leaderboard Space: MT-Bench column of the published leaderboard table1▇▇▇▇▇▇▇▇▇▇ 34%Apache-2.0
Hallucinations Leaderboard results (HF dataset hallucinations-leaderboard/results)1▇ 3%Apache-2.0
prior · pull toward the middle▇ 2%
weighted average= 28.2

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json