whatmodel.sh / GPT-4 Turbo (2024-04-09)

GPT-4 Turbo (2024-04-09)

+ contribute

GPT-4 Turbo (2024-04-09) is a closed-weights language model from OpenAI, released April 2024. Licence: proprietary.

Catalogue identity notes

Comparison model from the Epoch AI internal-evaluation CSVs (Epoch release date 2024-04-09). Dated API snapshot gpt-4-turbo-2024-04-09. OpenAI's deprecations page schedules its shutdown for October 23, 2026 (announced 2026-04-22).

55.4±0.4#255 of 719 LLMsfrom 49 benchmark results across 24 sources

$ how it compares

The LLMs ranked either side of GPT-4 Turbo (2024-04-09), out of 828 with a score, then more from OpenAI.

#	model	score	where	RAM	released
305	Nemotron Nano 9B v2 NVIDIA	56±1	api $0.10/M +local	~7.6GB	2025-08
306	Magistral Medium 1.0 (2506) Mistral	56±1	comparison only		2025-06
307	GPT-4 Turbo (2024-04-09) OpenAI	55±0	api $15.00/M		2024-04
308	Yi-Lightning 01.AI	55±1	api price unknown		2024-10
309	Qwen2.5 72B Instruct Alibaba	55±1	api $0.37/M +local	~30GB	2024-09
	more from OpenAI
1	GPT-6 Astra OpenAI	87±1	api $20.00/M		2026-09
3	GPT-6.1 Sol OpenAI	85±1	api $4.00/M		2026-09
7	GPT-6 Sol OpenAI	83±1	api $4.00/M		2026-09

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #168 of 300 * · 0.7% of its score · 2 tests

Beats o1 (2024-12-17), o3-mini, DeepSeek R1 0528 here, though they rank higher overall. Trails SmolLM3 3B, Jamba 1.5 Mini, Jamba 1.5 Large here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
BFCL v3	67.9% ✎	▁▁▁▁▁▃▅█▅▄▁▁ beats 103/114	medium · avg 52%	▇ 0.4%	blunt · at its level
BFCL (version not stated)	81.2% ✎	█·········██ beats 3/5	easy · avg 78%	▇ 0.3%	blunt

coding▁▁▁▃▆██▇▆▄▃▁ #187 of 458 · 24% of its score · 7 tests

Beats Gemini 2.5 Flash, Qwen3.5 4B, Qwen3 30B-A3B Instruct 2507 here, though they rank higher overall. Trails QwQ-32B-Preview, OpenThinker 32B, Command A Reasoning (08-2025) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
HumanEval	90.2% ✎	▁·▁▁▃▄▆▅█▃▃▁ beats 96/99	easy · avg 79%	▇▇▇ 7.2%	sharp
HumanEval+	86.6% ✎	▁·▁▁▁▃▄▄▄█▃▃ beats 96/99	easy · avg 74%	▇▇▇ 6.7%	sharp
CRUXEval (ZeroEval, 800 problems)	78.9% ✎	▁▂▆▃▆█▆▃▂·▁▁ beats 40/47	medium · avg 61%	▇▇ 3.9%	sharp
Aider code editing (133 Exercism Python exercises, two attempts)	63.9% ✎	▁·▁▂▄▁▆▄█▆▂▂ beats 33/62	medium · avg 54%	▇ 3.0%	at its level
LiveCodeBench official (Aug 1, 2024–May 1, 2025)	28.7% ✎	▃▃▅▃··▃▅█▅▅▃ beats 2/15	hard · avg 25%	▇ 1.4%	at its level
SciCode main problems (official leaderboard)	1.5% ✎	▃···█··▂·▃·▂ beats 8/17	very hard · avg 1.1%	▇ 0.9%	sharp · too hard for it
HumanEval (simple-evals) lab	88.2% ✎	▆█▆▆▂··▂···▂ beats 4/15	easy · avg 84%	▇ 0.6%	typical

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #186 of 393 · 15% of its score · 4 tests

Beats Seed-OSS 36B Instruct, Nemotron-Cascade-14B-Thinking, Granite 4.2 8B here, though they rank higher overall. Trails LFM2.5 1.2B Instruct, LFM2.5 1.2B Thinking, Gemma 3 12B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
AlpacaEval 2 length-controlled	55.0% ✎	▁▂▂▃▆▆█▆▄▄▁▁ beats 114/122	medium · avg 39%	▇▇▇ 7.9%	at its level
Arena-Hard (v0.1, GPT-4-1106 judge)	82.6% ✎	▂▁▁▅█▅▄▃▃▅▃▁ beats 55/60	medium · avg 52%	▇▇ 3.9%	sharp · at its level
Polyrating rating (bias-aware re-fit of Chatbot Arena votes, mid-2024)	1258.0 ✎	▂▁▂▁▃▇█▆▇▄▃▂ beats 91/95	avg 1201	▇ 1.8%	blunt
LMArena (text)	1324.5 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 161/379	avg 1268	▇ 1.8%	blunt

math▁▂▃▅▆▅▄█▄▁▁▁ #256 of 398 · 12% of its score · 7 tests

Trails DeepSeek R1 Distill Qwen 1.5B, Mixtral 8x7B v0.1 (base), Tulu 2 DPO 70B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
MATH level 5	46.7% ✎	▂▂▅▆▄█▆▃▅▂▃▄ beats 39/92	hard · avg 32%	▇▇ 4.4%	sharp · at its level
MATH Level 5 (ZeroEval, 721 problems)	46.5% ✎	▁▁▁▆█▂▇▇▄··▂ beats 23/38	hard · avg 34%	▇ 2.7%	sharp · at its level
OTIS mock AIME 2024–2025	6.7% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 39/193	very hard · avg 8.5%	▇ 1.3%	sharp · noisy
FrontierMath tiers 1–3 v2	0.7% ✎	▃··▁▁▃▄▄█▅▃▃ beats 3/82	very hard · avg 1.2%	▇ 1.2%	sharp · too hard for it
MATH (simple-evals, original test set, pre-o1 rows) lab	73.4% ✎	█····▄··▄··█ beats 3/5	medium · avg 63%	▇ 0.8%	at its level
MGSM (simple-evals) lab	89.6% ✎	▁····▁▅▁█▁▃▁ beats 8/16	easy · avg 82%	▇ 0.6%	blunt
MathVista (testmini)	47.5% ✎	▁··▁▄▄▇█▅▆█▄ beats 6/47	medium · avg 52%	▇ 0.5%	at its level · small run

vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #117 of 147 · 9.3% of its score · 12 tests

Beats DeepSeek V4 Flash Vision (Exp) here, though they rank higher overall. Trails Qwen3.5 2B, Qwen3-VL 4B Instruct, Gemma 4 E4B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
CharXiv descriptive questions	79.9% ✎	▄█▄▄▄·██·▄·█ beats 10/12	easy · avg 71%	▇ 1.6%	typical
CharXiv (reasoning questions)	37.1% ✎	▂▂▂█▃▃·▂▂··▂ beats 10/13	medium · avg 36%	▇ 1.2%	at its level
Arena vision (style controlled)	1112.0 ✎	▁▁▂▂▂▂▄▅▅▆██ beats 32/145	avg 1068	▇ 1.2%	blunt
MMMU val (VHELM, zero-shot exact match)	55.4% ✎	▄▄··▄▄█··▄▄█ beats 4/9	medium · avg 51%	▇ 1.1%	at its level
MathVista testmini (VHELM, zero-shot exact match)	49.3% ✎	▃·▅···▃▃▃·▃█ beats 4/9	medium · avg 43%	▇ 1.0%	at its level
CharXiv descriptive questions (test split)	79.8% ✎	█··▄··█··▄·▄ beats 5/6	easy · avg 71%	▇ 1.0%	typical
MMMU (val)	55.5% ✎	▁····▁▁▄█▇▆▆ beats 20/52	medium · avg 50%	▇ 0.7%	at its level · small run
CharXiv reasoning questions (test split)	33.8% ✎	▄··▄▄▄█····▄ beats 5/6	hard · avg 32%	▇ 0.7%	blunt · at its level

4 more tests

test	result	vs the others in its run	difficulty	counts	why
MathVista (testmini)	47.5% ✎	▁··▁▄▄▇█▅▆█▄ beats 6/47	medium · avg 52%	▇ 0.5%	at its level · small run
TextVQA	68.1% ✎	▁▁▁▄▅▁▃·▃▃▃█ beats 5/24	easy · avg 76%	▇ 0.1%	blunt · small run
ChartQA	62.3% ✎	▁··▁▁▁▃▁▂▅▅█ beats 4/44	easy · avg 79%	▇ 0.1%	blunt · noisy · small run
AI2D test	74.7% ✎	▁▁▅▇▃▇·▄·█·▃ beats 6/28	easy · avg 83%	▇ 0.0%	blunt · noisy · small run

other tests18 not in a task · 40% of its score

test	result	vs the others in its run	difficulty	counts	why
LiveBench 2024-07-26	52.9% ✎	▁▁▁▂▂▃▂█▆▃▅▃ beats 56/61	medium · avg 41%	▇▇▇ 7.9%	at its level
MMLU-Redux (ZeroEval, 2,778 questions)	85.3% ✎	▁·▁▁▄▄▄▅█▄▂▁ beats 43/53	easy · avg 77%	▇▇ 6.1%	typical
ZebraLogic (ZeroEval, 1,000 puzzles)	28.4% ✎	▁▁▃▆██▁▁·▁▁▁ beats 46/57	hard · avg 18%	▇▇ 4.0%	sharp · at its level
GPQA Diamond (simple-evals) lab	49.3% ✎	█▃▃█·▃▃▅▃▃▃▅ beats 4/16	hard · avg 40%	▇ 3.7%	sharp · at its level
GPQA Diamond	46.6% ✎	▁▁▁▁▂▄▆▄▅▇█▅ beats 55/208	hard · avg 40%	▇ 3.0%	sharp · at its level
BigCodeBench (Instruct)	48.0% ✎	▃▂·▆▆██▆▅▃▃▃ beats 32/33	medium · avg 41%	▇ 2.4%	at its level
CLIcK Korean (Daekeun Kim)	72.8% ✎	▆▄▄▄█▄▂·▄█·▄ beats 11/23	medium · avg 58%	▇ 2.3%	at its level
BigCodeBench (Complete)	58.2% ✎	▂▁▁▄▂▅▇█▆▂▂▃ beats 57/58	medium · avg 50%	▇ 2.1%	at its level

10 more tests

test	result	vs the others in its run	difficulty	counts	why
WildBench v2 WB-Score (task macro)	55.2212 WB-Score ✎	▁·▁▁·▃█▄▇▆▅▆ beats 47/52	avg 45.199 WB-Score	▇ 2.0%	sharp
HAE-RAE 1.0 (Daekeun Kim)	77.8% ✎	▄▄▄▂▂█▂▄▂█▂▄ beats 9/22	medium · avg 63%	▇ 2.0%	at its level
MMLU (simple-evals, zero-shot CoT) lab	86.7% ✎	▂▂·█▄▄▂·█▂·▂ beats 6/16	easy · avg 80%	▇ 1.3%	typical
KMMLU-HARD (Daekeun Kim, zero-shot)	30.6% ✎	█···▃▁▁▂▃·▆▂ beats 8/23	very hard · avg 29%	▇ 1.0%	sharp
Polish EQ-Bench v2 (SpeakLeash)	77.8% ✎	▁··▁▁▂▂▆▅▆█▆ beats 69/70	medium · avg 64%	▇ 0.8%	at its level
Open Persian LLM Leaderboard v2 average (21 tasks)	70.0% ✎	▂▄▂▃█▄▃▆▃▅▂▂ beats 73/81	medium · avg 58%	▇ 0.7%	blunt · at its level
DROP F1, 3-shot (simple-evals) lab	86.0% ✎	▂▂▄▆▆█·▃···▄ beats 14/16	easy · avg 78%	▇ 0.2%	blunt
Chess Puzzles (Epoch)	6.0% ✎	█▃▁▁▁▂▁▄▂▂▁▁ beats 46/135	very hard · avg 5.9%	▇ 0.2%	sharp · too hard for it · noisy
SimpleQA (simple-evals) lab	24.2% ✎	▄·▁▁▁▃··█▁·▁ beats 6/14	hard · avg 18%	▇ 0.1%	noisy
Vectara hallucination leaderboard, 2025 dataset (HHEM-2.3, factual consistency rate)	98.3% ✎	▁▁▂▃▄▄█▆▃▃▂▁ beats 127/140	very easy · avg 94%	▇ 0.1%	blunt · maxed out · noisy

$ where to run it

Through an API

api · $15.00/M · $10 in / $30 out per 1M tokens · 128k context · developers.openai.com ↗ · price sourceUSD per 1M tokens, OpenAI API standard processing, model page 'gpt-4-turbo' whose default snapshot is gpt-4-turbo-2024-04-09: input $10, output $30; cached-input, Batch and other tiers excluded. Currency: USD. OpenAI official model page, observed 2026-09-28.

$ how the score was worked out · WildEval ZeroEval published r… 17%, EvalPlus / EvalPerf published… 14%, Epoch AI — internal evaluatio… 10% and 21 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

source	results	share of the score	licence
WildEval ZeroEval published results	4	▇▇▇▇▇ 17%	Apache-2.0
EvalPlus / EvalPerf published evaluation runs	2	▇▇▇▇ 14%	Apache-2.0
Epoch AI — internal evaluations	5	▇▇▇ 10%	CC BY 4.0
AlpacaEval 1.0 and 2.0 leaderboards (tatsu-lab/alpaca_eval repository)	1	▇▇ 8%	Apache-2.0
LiveBench model judgments (HF livebench/model_judgment, 2024-10-22 revision) with the LiveBench question datasets	1	▇▇ 8%	Apache-2.0 (stated in the LiveBench datasheet)
OpenAI simple-evals README benchmark results (OpenAI's own models) · lab-reported	7	▇▇ 7%	MIT
Daekeun Kim Korean LLM benchmark runs	3	▇▇ 5%	MIT
BigCodeBench paper (Zhuo et al., ICLR 2025), arXiv 2406.15877v4	2	▇ 4%	CC BY 4.0
Arena-Hard-Auto v0.1 leaderboard (repository README, pre-v2)	1	▇ 4%	Apache-2.0
Aider code editing benchmark runs (old edit leaderboard)	1	▇ 3%	Apache-2.0
LMArena leaderboard dataset	2	▇ 3%	CC BY 4.0
CharXiv leaderboard (princeton-nlp/CharXiv README)	2	▇ 3%	Apache-2.0
VHELM: A Holistic Evaluation of Vision Language Models (Stanford CRFM), arXiv 2410.07112v2	2	▇ 2%	CC BY 4.0
WildBench v2 leaderboard data (Ai2)	1	▇ 2%	Apache-2.0
Polyrating: A Cost-Effective and Bias-Aware Rating System for LLM Evaluation (arXiv:2409.00696v3)	1	▇ 2%	CC-BY-4.0
CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMs (arXiv:2406.18521v1)	2	▇ 2%	CC BY-SA 4.0
LiveCodeBench official website results (2025-08-01)	1	▇ 1%	CC-BY-SA-4.0
Phi-3 Technical Report (Microsoft, arXiv 2404.14219v3)	5	▇ 1%	CC BY 4.0
SciCode leaderboard (repository README)	1	▇ 1%	Apache-2.0
SpeakLeash Polish EQ-Bench leaderboard	1	▇ 1%	Apache-2.0
Open Persian LLM Leaderboard v2 (Part DP AI and AUT NLP Lab)	1	▇ 1%	Apache-2.0
Berkeley Function Calling Leaderboard result archive	1	▇ 0%	Apache-2.0
ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools (Team GLM, arXiv 2406.12793v2)	1	▇ 0%	CC BY 4.0
Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository)	1	▇ 0%	Apache-2.0
prior · pull toward the middle		▇ 0%
weighted average		= 55.4