GPT-4 (Mar 2023)
GPT-4 (Mar 2023) (OpenAI) was released on Mar 14, 2023. In the Epoch AI capability ranking (ECI), it ranks #178 out of 267 models, with 125.88. Against models of the same generation, its best result is on WinoGrande: 87.5%, #1 of 17.
Where GPT-4 (Mar 2023) stands out and where it falls behind, by the numbers
GPT-4 (Mar 2023) has published results on 15 tests compiled by Epoch AI. The yardstick here is its generation: its position among the models released between Jun 17, 2022 and Jun 12, 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.
It is not in the bottom half of its generation on any test with 6 or more models compared.
See all 15 tests, one by one
| Test | Score | In its generation | Overall | Test record |
|---|---|---|---|---|
| GPQA Diamond | 35.7% | #1 / 1 | #154 / 173 | 96.4% Gemini 4 Argon |
| AIME 2024 and 2025 (OTIS) | 1.1% | #1 / 1 | #162 / 168 | 100.0% GPT-5.5 |
| Dtbench | 62.7% | #1 / 1 | #105 / 148 | 98.9% Claude Opus 5.5 |
| Lmca | 17.06 | #1 / 1 | #98 / 121 | 68.23 Claude Opus 5.5 |
| WeirdML | 12.4% | #1 / 1 | #103 / 112 | 93.6% GPT-6 Astra |
| MMLU | 86.4% | #1 / 13 | #5 / 92 | 88.1% GPT-4o (May 2024) |
| Game puzzles | 12.0% | #1 / 1 | #58 / 77 | 84.0% GPT-6 Astra |
| MATH level 5 | 23.0% | #1 / 1 | #56 / 73 | 98.1% GPT-5 |
| ForecastBench | 57.80 | #1 / 1 | #47 / 62 | 62.50 o3 |
| WinoGrande | 87.5% | #1 / 17 | #3 / 61 | 89.2% Llama 3.1-405B |
| GSM8K | 92.0% | #1 / 10 | #1 / 54 | 92.0% GPT-4 (Mar 2023) |
| HellaSwag | 95.3% | #1 / 17 | #1 / 51 | 95.3% GPT-4 (Mar 2023) |
| BIG-Bench Hard | 75.1% | #1 / 10 | #8 / 41 | 89.2% Gemini 1.5 Pro (Sept 2024) |
| TriviaQA | 84.8% | #3 / 11 | #6 / 32 | 87.6% Llama 2-70B |
| METR: time horizon | 5.36 | #1 / 1 | #26 / 29 | 719 Claude Opus 4.6 |
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 GPT-4 (Mar 2023) 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-4 (Mar 2023) | OpenAI | 125.88 | n/a | n/a | Mar 14, 2023 |
| Claude Instant | Anthropic | 120.31 | n/a | n/a | Aug 9, 2023 |
| Claude 2 | Anthropic | 120.15 | n/a | n/a | Jul 11, 2023 |
| 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 |
| PaLM 2-L | 115.07 | n/a | n/a | May 17, 2023 | |
| Llama 2-70B | Meta AI | 113.84 | n/a | n/a | Jul 18, 2023 |
GPT-4 (Mar 2023) or Claude 2?
Claude 2 is the model from another lab closest to GPT-4 (Mar 2023) in capability. Across the 6 tests both took, this is the score:
| Test | GPT-4 (Mar 2023) | Claude 2 |
|---|---|---|
| AIME 2024 and 2025 (OTIS) | 1.1% | 2.5% |
| MATH level 5 | 23.0% | 11.7% |
| Dtbench | 62.7% | 51.9% |
| MMLU | 86.4% | 78.5% |
| TriviaQA | 84.8% | 87.5% |
| GPQA Diamond | 35.7% | 34.7% |
Unweighted score: each test counts one point, whether math or coding. The dot marks who did better on each row.
Other models from OpenAI
The models from the same lab closest in time, newest first.
| Model | Release | ECI | ELO | LiveBench | Access |
|---|---|---|---|---|---|
| GPT-4.5 | Feb 27, 2025 | 136.73 | n/a | n/a | API access |
| o3-mini | Jan 31, 2025 | 140.33 | 1,379 | n/a | API access |
| o1 | Dec 17, 2024 | 141.84 | 1,383 | n/a | API access |
| o1-mini | Sep 12, 2024 | 135.82 | 1,362 | n/a | API access |
| GPT-4o mini | Jul 18, 2024 | 126.58 | n/a | n/a | API access |
| GPT-4o (May 2024) | May 13, 2024 | 129.05 | n/a | n/a | API access |
| GPT-4 Turbo (Apr 2024) | Apr 9, 2024 | 127.25 | n/a | n/a | API access |
| GPT-3.5 Turbo (Nov 2023) | Nov 6, 2023 | 118.58 | n/a | n/a | API access |
| GPT-4 (Mar 2023) | Mar 14, 2023 | 125.88 | n/a | n/a | API access |
Have you used GPT-4 (Mar 2023)? 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 GPT-4 (Mar 2023)
Where does GPT-4 (Mar 2023) rank among AI models?
In the ROO3 AI Benchmark, ordered by the Epoch AI ECI, GPT-4 (Mar 2023) ranks #178 out of 267 models, with an ECI of 125.88. 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 GPT-4 (Mar 2023) better than Claude 2?
Across the 6 tests both models took and Epoch AI compiled, GPT-4 (Mar 2023) comes out ahead on 4, behind on 2 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 GPT-4 (Mar 2023) 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 GPT-4 (Mar 2023) are: WinoGrande (#1 of 17); HellaSwag (#1 of 17); MMLU (#1 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