Stable Beluga 2
Stable Beluga 2 (Stability AI) was released on Jul 20, 2023. In the Epoch AI capability ranking (ECI), it ranks #205 out of 267 models, with 117.16. Against models of the same generation, its best result is on ARC (AI2): 86.1%, #2 of 34.
Where Stable Beluga 2 stands out and where it falls behind, by the numbers
Stable Beluga 2 has published results on 8 tests compiled by Epoch AI. The yardstick here is its generation: its position among the models released between Oct 23, 2022 and Oct 18, 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 8 tests, one by one
| Test | Score | In its generation | Overall | Test record |
|---|---|---|---|---|
| MMLU | 68.6% | #6 / 31 | #46 / 92 | 88.1% GPT-4o (May 2024) |
| ARC (AI2) | 86.1% | #2 / 34 | #10 / 65 | 95.3% DeepSeek-V3 (Mar 2025) |
| GSM8K | 69.6% | #3 / 26 (tied with 1) | #14 / 54 | 92.0% GPT-4 (Mar 2023) |
| HellaSwag | 84.1% | #7 / 30 | #12 / 51 | 95.3% GPT-4 (Mar 2023) |
| PIQA | 83.3% | #2 / 29 | #10 / 43 | 88.7% GPT-4o mini |
| BoolQ | 89.4% | #2 / 33 | #2 / 42 | 90.9% PaLM 2-L |
| BIG-Bench Hard | 69.3% | #2 / 24 | #11 / 41 | 89.2% Gemini 1.5 Pro (Sept 2024) |
| LAMBADA | 71.3% | #14 / 19 | #14 / 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 Stable Beluga 2 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 |
|---|---|---|---|---|---|
| GPT-3.5 Turbo (Nov 2023) | OpenAI | 118.58 | n/a | n/a | Nov 6, 2023 |
| Mixtral 8x7B | Mistral AI | 118.50 | n/a | n/a | Dec 11, 2023 |
| Claude 3 Haiku | Anthropic | 118.39 | 1,206 | n/a | Mar 7, 2024 |
| Yi-34B | 01.AI | 117.43 | n/a | n/a | Nov 2, 2023 |
| Stable Beluga 2 | Stability AI | 117.16 | n/a | n/a | Jul 20, 2023 |
| Gemini 1.0 Pro | Google DeepMind | 117.07 | n/a | n/a | Dec 13, 2023 |
| PaLM 2-L | 115.07 | n/a | n/a | May 17, 2023 |
Stable Beluga 2 or Mixtral 8x7B?
Mixtral 8x7B is the model from another lab closest to Stable Beluga 2 in capability. Across the 5 tests both took, this is the score:
| Test | Stable Beluga 2 | Mixtral 8x7B |
|---|---|---|
| GSM8K | 69.6% | 74.4% |
| HellaSwag | 84.1% | 86.7% |
| MMLU | 68.6% | 70.6% |
| ARC (AI2) | 86.1% | 87.3% |
| PIQA | 83.3% | 83.6% |
Unweighted score: each test counts one point, whether math or coding. The dot marks who did better on each row.
Have you used Stable Beluga 2? 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 Stable Beluga 2
Where does Stable Beluga 2 rank among AI models?
In the ROO3 AI Benchmark, ordered by the Epoch AI ECI, Stable Beluga 2 ranks #205 out of 267 models, with an ECI of 117.16. 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 Stable Beluga 2 better than Mixtral 8x7B?
Across the 5 tests both models took and Epoch AI compiled, Stable Beluga 2 comes out ahead on 0, behind on 5 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 Stable Beluga 2 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 Stable Beluga 2 are: ARC (AI2) (#2 of 34); BoolQ (#2 of 33); PIQA (#2 of 29). 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