whatmodel.sh / GPT-3.5 Turbo (0613)
GPT-3.5 Turbo (0613)
↗ share
+ contribute
GPT-3.5 Turbo (0613) is a closed-weights language model from OpenAI, released June 2023. Licence: proprietary.
Catalogue identity notes
Comparison-only. Dated API snapshot gpt-3.5-turbo-0613 (4K context; not gpt-3.5-turbo-16k-0613). OpenAI's deprecations page lists its shutdown on 2024-09-13, so no current API. Release month from the snapshot date in the model name.
$ how it compares
The LLMs ranked either side of GPT-3.5 Turbo (0613), out of 828 with a score, then more from OpenAI.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 512 |  Gemini 1.0 Pro (001) | 46±2 | comparison only | ||
| 513 |  Mistral Medium (2312) | 46±1 | comparison only | ||
| 514 | GPT-3.5 Turbo (0613) | 46±1 | comparison only | ||
| 515 |  Granite 3.3 8B Instruct | 46±1 | local | ||
| 516 |  Tulu 2 70B | 46±2 | local | ||
| more from OpenAI | |||||
| 1 | GPT-6 Astra | 87±1 | api $20.00/M | ||
| 3 | GPT-6.1 Sol | 85±1 | api $4.00/M | ||
| 7 | GPT-6 Sol | 83±1 | api $4.00/M | ||
$ what it's good at
coding▁▁▁▃▅▇█▇▅▄▃▁ #324 of 524 · 18% of its score · 4 tests
Beats Qwen3-VL 32B Instruct, Solar Open 100B (102B-A12B), GLM-4.7-Flash here, though they rank higher overall. Trails DeepSeek LLM 67B Chat, Mistral NeMo Instruct (2407), QwQ-32B-Preview here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Aider code editing (133 Exercism Python exercises, two attempts) | 50.4% ✎ | ▁·▁▂▄▁▆▄█▆▂▂ beats 14/62 | 11% |
| Tulu 2: CodexEval pass@10 | 88.0% ✎ | ▃▃▃·▆▃▃█▃▂▂▅ beats 24/25 | 3.3% |
| CRUXEval-I | 49.0% ✎ | ▁··▁▅▂█▇▃··▁ beats 21/22 | 2.3% |
| CRUXEval-O | 49.4% ✎ | ▁·▂▃▁█▁▃···▁ beats 20/22 | 2.2% |
chat & writing▁▁▁▃▅▄▆█▇▇▇▃ #352 of 463 · 45% of its score · 4 tests
Beats Ministral 3 14B Reasoning 2512, Gemma 4 E4B, OpenThinker3 7B here, though they rank higher overall. Trails Flan-T5 XXL, LFM2 700M, EXAONE 3.5 2.4B Instruct here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| AlpacaEval 2 length-controlled | 22.4% ✎ | ▁▂▂▃▆▆█▆▄▄▁▁ beats 57/122 | 20% |
| Arena-Hard (v0.1, GPT-4-1106 judge) | 24.8% ✎ | ▂▁▁▅█▅▄▃▃▅▃▁ beats 24/60 | 14% |
| Polyrating rating (bias-aware re-fit of Chatbot Arena votes, mid-2024) | 1161.0 ✎ | ▂▁▂▁▃▇█▆▇▄▂▂ beats 70/95 | 5.7% |
| MT-Bench (GPT-4 single-answer grading, LMSYS runs) | 8.3900 MT-Bench score (1-10) ✎ | ▆▂▂▆▆▄██▂▄▄▆ beats 25/28 | 5.0% |
math▁▂▃▅▇▅▅█▄▁▁▁ #250 of 462 * · 2.9% of its score · 1 test
Beats GPT-4o (2024-08-06), Command A (03-2025), GPT-4o (2024-11-20) here, though they rank higher overall. Trails Olmo 3 7B Instruct here, though they rank lower overall.
other tests7 not in a task · 33% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| BIG-Bench Hard | 70.8% ✎ | ▁·····▂█▂▅·▁ beats 24/25 | 9.3% |
| CLIcK Korean (Daekeun Kim) | 51.0% ✎ | ▆▄▄▄█▄▂·▄█·▄ beats 3/23 | 8.4% |
| HAE-RAE 1.0 (Daekeun Kim) | 52.7% ✎ | ▄▄▄▂▂█▂▄▂█▂▄ beats 2/22 | 7.6% |
| MMLU | 65.7% ✎ | ▁▁▃▃▄█▄▁█··▁ beats 21/25 | 3.4% |
| EQ-Bench v1 (60 questions, paper run) | 49.2% ✎ | ▂··▄·▂▆▄▂█·▂ beats 10/14 | 2.6% |
| Tulu 2: TyDiQA gold passage, 1-shot F1 | 51.2% ✎ | ▂··▁▃▂▂▂█▅▁▂ beats 13/25 | 1.0% |
| KMMLU-HARD (Daekeun Kim, zero-shot) | 21.0% ✎ | █···▁▁▁▂▃·▆▂ beats 0/23 | 1.0% |
$ where to run it
No verified API or download recorded yet · add one
$ how the score was worked out · AlpacaEval 1.0 and 2.0 leader… 20%, Tulu 2 — base and instruction… 20%, Daekeun Kim Korean LLM benchm… 17% and 6 more
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
$ improve this page
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