Falcon-40B
Falcon-40B (Technology Innovation Institute) was released on May 25, 2023. In the Epoch AI capability ranking (ECI), it ranks #237 out of 267 models, with 104.54. Against models of the same generation, its best result is on PIQA: 83.0%, #2 of 23.
Where Falcon-40B stands out and where it falls behind, by the numbers
Falcon-40B has published results on 11 tests compiled by Epoch AI. The yardstick here is its generation: its position among the models released between Aug 28, 2022 and Aug 23, 2023 that took the same test. Against the whole archive, which includes 2023 models, a recent release lands near the top of almost everything, and that says nothing.
Where it stands out
Tests where it sits in the top third of its own generation.
Where it falls behind
Tests where it sits in the bottom half of its own generation.
See all 11 tests, one by one
| Test | Score | In its generation | Overall | Test record |
|---|---|---|---|---|
| MMLU | 56.9% | #9 / 24 | #63 / 92 | 88.1% GPT-4o (May 2024) |
| ARC (AI2) | 61.9% | #9 / 26 | #28 / 65 | 95.3% DeepSeek-V3 (Mar 2025) |
| WinoGrande | 76.9% | #7 / 23 | #22 / 61 | 89.2% Llama 3.1-405B |
| GSM8K | 33.8% | #10 / 19 | #38 / 54 | 92.0% GPT-4 (Mar 2023) |
| HellaSwag | 85.3% | #4 / 24 | #9 / 51 | 95.3% GPT-4 (Mar 2023) |
| PIQA | 83.0% | #2 / 23 | #11 / 43 | 88.7% GPT-4o mini |
| BoolQ | 83.1% | #11 / 25 | #20 / 42 | 90.9% PaLM 2-L |
| BIG-Bench Hard | 37.1% | #12 / 18 | #32 / 41 | 89.2% Gemini 1.5 Pro (Sept 2024) |
| OpenBookQA | 56.6% | #9 / 21 | #19 / 32 | 88.0% phi-3-mini 3.8B |
| TriviaQA | 79.9% | #9 / 18 | #14 / 32 | 87.6% Llama 2-70B |
| LAMBADA | 77.3% | #3 / 13 | #4 / 20 | 79.8% Falcon-180B |
For each test, the model's best effort level counts, as in the ranking. Source of the scores: Epoch AI, under CC BY 4.0.
What to compare Falcon-40B with
The models from other labs closest in capability and released in the same period, at most two per lab. Click any of them to see its page.
| Model | Lab | ECI | ELO | LiveBench | Release |
|---|---|---|---|---|---|
| PaLM 2-S | 106.26 | n/a | n/a | May 17, 2023 | |
| Llama 2-13B | Meta AI | 106.25 | n/a | n/a | Jul 18, 2023 |
| Llama 2-34B | Meta AI | 105.29 | n/a | n/a | Jul 18, 2023 |
| Yi 6B | 01.AI | 104.86 | n/a | n/a | Nov 22, 2023 |
| Falcon-40B | Technology Innovation Institute | 104.54 | n/a | n/a | May 25, 2023 |
| Baichuan2-13B | Baichuan | 103.29 | n/a | n/a | Sep 6, 2023 |
| internlm-7b | 102.97 | n/a | n/a | Jul 5, 2023 |
Falcon-40B or Yi 6B?
Yi 6B is the model from another lab closest to Falcon-40B in capability. Across the 6 tests both took, this is the score:
| Test | Falcon-40B | Yi 6B |
|---|---|---|
| GSM8K | 33.8% | 44.9% |
| BIG-Bench Hard | 37.1% | 47.2% |
| ARC (AI2) | 61.9% | 50.3% |
| HellaSwag | 85.3% | 74.4% |
| MMLU | 56.9% | 64.0% |
| WinoGrande | 76.9% | 71.3% |
Unweighted score: each test counts one point, whether math or coding. The dot marks who did better on each row.
Other models from Technology Innovation Institute
The models from the same lab closest in time, newest first.
| Model | Release | ECI | ELO | LiveBench | Access |
|---|---|---|---|---|---|
| Falcon 2 11B | May 9, 2024 | 109.63 | n/a | n/a | Open weights (restricted use) |
| Falcon-180B | Sep 6, 2023 | 112.19 | n/a | n/a | Open weights (restricted use) |
| Falcon-40B | May 25, 2023 | 104.54 | n/a | n/a | Open weights (unrestricted) |
| Falcon-7B | Apr 24, 2023 | 95.23 | n/a | n/a | Open weights (unrestricted) |
Have you used Falcon-40B? Rate it
Users' opinion, not measurement: the rating is kept apart from the ranking and does not move anyone's position. One vote per browser, no sign-up, and you can change your rating later.
What people ask about Falcon-40B
Where does Falcon-40B rank among AI models?
In the ROO3 AI Benchmark, ordered by the Epoch AI ECI, Falcon-40B ranks #237 out of 267 models, with an ECI of 104.54. The ECI is a composite index that combines several hard benchmarks into a single scale. The position changes when a new model comes out, and the page is refreshed six times a day.
Is Falcon-40B better than Yi 6B?
Across the 6 tests both models took and Epoch AI compiled, Falcon-40B comes out ahead on 3, behind on 3 and tied on 0. It is a test-by-test count with no weighting: a win in math counts the same as a win in coding. To choose between the two, look at the tests that resemble your own use.
What is Falcon-40B best at?
Against models of the same generation (released up to 9 months before and up to 3 months after it), the best relative results of Falcon-40B are: PIQA (#2 of 23); HellaSwag (#4 of 24); LAMBADA (#3 of 13). The yardstick is its position among the models of that period that took the same test.
Where do the numbers on this page come from?
From the same three sources as the ranking, all openly licensed: Epoch AI (ECI and individual tests, CC BY 4.0), Arena (human-preference ELO, CC BY 4.0) and LiveBench (overall average, CC BY-SA 4.0). ROO3 does not measure any model. When the page shows a figure published by the maker itself, it appears in the written analysis, labeled as the maker's, and never enters the ranking.
Where each number comes from
Epoch AI, "AI Benchmarking Hub". Published at epoch.ai. Data under the CC BY 4.0 licence, reordered by roo3.co.
The Arena (LMArena) leaderboard, from the public lmarena-ai/leaderboard-dataset on Hugging Face, under the CC BY 4.0 licence. We are not affiliated with the Arena.
LiveBench, by the LiveBench team, published at livebench.ai under the CC BY-SA 4.0 license. The overall average is recomputed by roo3.co with the same formula as the website (average of the categories), and those numbers stay under CC BY-SA 4.0. We are not affiliated with LiveBench.
← See the full ranking