whatmodel.sh / Qwen3 30B-A3B Thinking 2507

Qwen3 30B-A3B Thinking 2507

+ contribute

Qwen3 30B-A3B Thinking 2507 is an open-weights language model from Alibaba, released July 2025. Its published size is 30.5B parameters (3.3B active per token) and a 262k-token context window. Licence: apache-2.0.

60.6±0.5#187 of 719 LLMsfrom 93 benchmark results across 14 sources

$ how it compares

The LLMs ranked either side of Qwen3 30B-A3B Thinking 2507, out of 828 with a score, then more from Alibaba.

#	model	score	where	RAM	released
223	o1-mini OpenAI	61±1	comparison only		2024-09
224	Gemini 2.0 Pro Experimental (02-05) Google	61±1	comparison only		2025-02
225	Qwen3 30B-A3B Thinking 2507 Alibaba	61±1	api $0.75/M +local	~20GB	2025-07
226	Qwen3.5 4B Alibaba	61±1	local	~4.3GB	2026-03
227	Nemotron-Cascade-14B-Thinking NVIDIA	61±1	local	~10GB	2025-12
	more from Alibaba
18	Qwen3.8 Max (2.4T-A95B) Alibaba	78±1	api $3.00/M +local	~380GB	2026-08
34	Qwen3.8 Max (0902) Alibaba	76±1	api $3.00/M		2026-09
37	Qwen3.8-Flash-Next Alibaba	75±1	local	~59GB	2026-08

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #174 of 300 · 21% of its score · 10 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
τ²-bench Telecom lab +3	26.3% ✎	▃·█········█ beats 1/6	very hard · avg 10%	▇▇▇ 8.6%	sharp · at its level · several runs
τ-bench Retail lab +2	67.8% ✎	▁····▃·▂▇█▂▆ beats 18/28	medium · avg 36%	▇ 3.4%	at its level · several runs
τ-bench Airline lab +2	48.0% ✎	▁▁▁·▅▄▁█▇▁▁▇ beats 18/28	hard · avg 25%	▇ 2.4%	at its level · small run · several runs
τ²-bench Retail lab +1	58.8% ✎	▁····▂▁▃▆█▃▁ beats 10/44	medium · avg 37%	▇ 1.9%	at its level · several runs
τ²-bench Airline lab +1	58.0% ✎	▁▁·▂▁▃▂▁█▅▂▃ beats 24/42	hard · avg 32%	▇ 1.3%	at its level · several runs
SWE-bench Verified	31.0% ✎	▁▁▁▁▁▁▃▄█▆▁▁ beats 19/131	hard · avg 17%	▇ 1.1%	sharp · at its level · small run
BFCL v3 lab +1	72.4% ✎	▁▁▁▁▁▃▅█▆▅▄▁ beats 161/170	medium · avg 59%	▇ 0.6%	blunt · at its level · several runs
τ²-bench aggregate lab	41.9% ✎	▁▁▁·▃▂▄▃▃▅█▂ beats 3/37	medium · avg 37%	▇ 0.5%	at its level · noisy

2 more tests

test	result	vs the others in its run	difficulty	counts	why
Multi-SWE-bench (scaffold per note)	9.5% ✎	█··█·······█ beats 1/2	very hard · avg 5.7%	▇ 0.5%	small run
BFCL V4 (December 2025 archive) lab	42.4% ✎	▁▁▁▂▇▆▆▆█▂▄▂ beats 55/111	medium · avg 36%	▇ 0.5%	at its level

coding▁▁▁▃▆██▇▆▄▃▁ #182 of 458 · 4.5% of its score · 3 tests

Beats Gemini 2.5 Flash here, though they rank higher overall. Trails QwQ-32B-Preview, OpenThinker 32B, Ministral 3 8B Reasoning 2512 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LiveCodeBench v6 (Feb–May 2025) +2	60.3% ✎	▁▁▃▁█▅▄▇▅▅▃█ beats 17/38	medium · avg 40%	▇ 2.9%	at its level · several runs
SWE-bench Verified	31.0% ✎	▁▁▁▁▁▁▃▄█▆▁▁ beats 19/131	hard · avg 17%	▇ 1.1%	sharp · at its level · small run
Multi-SWE-bench (scaffold per note)	9.5% ✎	█··█·······█ beats 1/2	very hard · avg 5.7%	▇ 0.5%	small run

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #146 of 393 · 5.3% of its score · 2 tests

Beats INTELLECT-3, Nemotron 3 Super 120B-A12B, MiniMax-M2 here, though they rank higher overall. Trails LFM2.5 1.2B Thinking, LFM2.5 8B-A1B, LFM2.5 2.6B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena-Hard v2 aggregate lab +1	56.0% ✎	▁▁··▄▃▁▃█▁▁▇ beats 8/22	hard · avg 28%	▇▇ 3.9%	sharp · at its level · several runs
IFEval (aggregation unspecified) lab +2	88.9% ✎	▁▁▃▁▂▄▄▆█▃▂▁ beats 104/141	easy · avg 78%	▇ 1.4%	several runs

math▁▂▃▅▆▅▄█▄▁▁▁ #155 of 398 · 20% of its score · 10 tests

Beats MiMo-V2.5 (310B-A15B), MiniMax-M2.7, MiniMax-M2.1 here, though they rank higher overall. Trails Qwen3-VL 4B Thinking, Granite 4.2 3B, DeepSeek R1 0528 Qwen3 8B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
HMMT 2025 (edition unspecified) lab +1	71.4% ✎	▃█▇▅▇▅▁·▁··▁ beats 19/32	hard · avg 33%	▇▇ 4.6%	sharp · at its level · several runs
AIME 2025 +1	81.3% ✎	▁·▁▁▁▂▅▆█▇▅▁ beats 100/133	medium · avg 55%	▇ 3.2%	sharp · small run · several runs
OTIS mock AIME 2024–2025	70.3% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 101/193	very hard · avg 8.5%	▇ 2.1%	sharp · at its level · noisy
MathArena ArXivMath January 2026 (23 problems)	39.1% ✎	▂··▄▂▄█▆▆▄▄█ beats 2/23	hard · avg 22%	▇ 1.9%	at its level
MathArena ArXivMath February 2026 (32 problems)	22.7% ✎	▁▁▃▁█▇▄▃▁▃·▃ beats 2/24	very hard · avg 9.1%	▇ 1.8%	typical
PolyMATH lab +1	52.6% ✎	▁···▃▁·▇█▄▄▄ beats 14/23	hard · avg 32%	▇ 1.6%	at its level · several runs
AIME 2024	87.7% ✎	▁▁▁▂▁▃▁▅███▃ beats 65/85	easy · avg 68%	▇ 1.6%	sharp
HMMT February 2025 lab	63.1% ✎	▁▁▁▃▁▅▅▃██▁▄ beats 7/31	medium · avg 48%	▇ 1.2%	at its level

2 more tests

test	result	vs the others in its run	difficulty	counts	why
MathArena ArXivMath December 2025 (17 problems)	27.9% ✎	▃·▃▃█▁·▄·▁▁▁ beats 1/18	hard · avg 19%	▇ 1.1%	at its level
HMMT November 2025 lab	73.8% ✎	▁▁····▃▅▃█▇▁ beats 2/23	medium · avg 61%	▇ 1.0%	typical

other tests35 not in a task · 51% of its score

test	result	vs the others in its run	difficulty	counts	why
GPQA Diamond +2	70.1% ✎	▁▁▁▁▂▄▆▄▅▇█▅ beats 105/208	hard · avg 40%	▇▇▇▇ 9.1%	sharp · at its level · several runs
GPQA (subset unspecified) lab +1	73.4% ✎	▂▁▂▂▄▃▆▃▅█▄▃ beats 83/104	hard · avg 49%	▇▇ 4.2%	sharp · at its level · several runs
MMLU-Pro +2	81.9% ✎	▁▁·▁▁▁▃▂▄▅█▁ beats 142/188	medium · avg 61%	▇ 3.5%	several runs
MMLU-ProX lab +3	76.4% ✎	▁·▁▁▁▂▁▃▂█▅▄ beats 15/35	medium · avg 60%	▇ 2.7%	several runs
SuperGPQA lab +2	56.8% ✎	▂▁▂▂▅▆▃▇▇▇█▃ beats 30/46	medium · avg 39%	▇ 2.2%	at its level · several runs
INCLUDE lab +1	74.4% ✎	▂····▂▂▄▄▄█▄ beats 21/36	medium · avg 58%	▇ 2.2%	at its level · several runs
MMMLU lab +1	78.4% ✎	▂▁▁▂▃▂▄▄▅██▃ beats 29/46	medium · avg 63%	▇ 2.1%	at its level · several runs
MMLU-Redux lab +1	91.4% ✎	▁▁▃▁▂▃▄▅▇▅█▄ beats 42/63	easy · avg 82%	▇ 2.0%	several runs

27 more tests

test	result	vs the others in its run	difficulty	counts	why
OJBench lab +1	25.1% ✎	▁···▁▃·▅▄▅██ beats 13/26	very hard · avg 13%	▇ 1.9%	at its level · several runs
Medmarks-V weighted mean win rate (61 models, 71 configurations)	55.9% ✎	▂▃▅▂▆▃█▅█▅▃▆ beats 25/34	medium · avg 42%	▇ 1.5%	blunt · at its level
C-Eval lab	87.4% ✎	▁···▁▁▂▁▂▂▃█ beats 18/33	medium · avg 73%	▇ 1.4%	typical
MMLU	86.9% ✎	▁▁▁▁▁▂▃▅█▃▆▃ beats 167/183	medium · avg 72%	▇ 1.4%	typical
WritingBench lab +1	85.0% ✎	▁·▁▂▁··▅█▂▂▁ beats 11/20	easy · avg 70%	▇ 1.3%	several runs
LongBench v2 lab	44.8% ✎	▂·▁·▂▂▁▃▄▁█▄ beats 13/28	very hard · avg 35%	▇ 1.3%	at its level
LiveBench 2024-11-25 lab +1	76.8% ✎	▄··▂·▂·▂█▆▂▂ beats 10/13	medium · avg 60%	▇ 1.2%	at its level · several runs
KINA (Knowledge Index of Noah's Ark) accuracy, 899 ten-option items	27.0% ✎	▁▁▂▁▃▄█▆▄▅▂▃ beats 13/39	very hard · avg 22%	▇ 1.2%	blunt
BRIDGE clinical text, zero-shot average (87 tasks)	41.8% ✎	▁▁▁▁▁▂▂▃▄▆█▃ beats 86/97	hard · avg 32%	▇ 1.1%	blunt · at its level
Humanity's Last Exam	8.7% ✎	▁▁▁▃▆▆▇██▆▁▃ beats 15/104	very hard · avg 4.2%	▇ 1.1%	typical
Multi-IF (aggregation unspecified) lab +1	76.4% ✎	▁▁▁▁▁▁▃▄█▄▁▂ beats 22/41	medium · avg 63%	▇ 1.1%	blunt · several runs
Global PIQA lab	80.2% ✎	▃▃··▃··▃▃▅█▅ beats 4/11	medium · avg 71%	▇ 0.9%	at its level
MAXIFE lab	77.4% ✎	▂·▂·▂···▆·▄█ beats 3/11	medium · avg 63%	▇ 0.9%	typical
MultiChallenge lab	46.5% ✎	▁·▁·▁▃▆▅▃▁█▅ beats 12/29	hard · avg 30%	▇ 0.8%	at its level
VitaBench lab	14.1% ✎	▄··▃▄▆▆█▄▄▄▂ beats 2/21	very hard · avg 10%	▇ 0.8%	typical
IFBench (aggregation unspecified) lab	51.5% ✎	▃▇▂▄▅▅▃▄▄▆█▇ beats 37/80	medium · avg 39%	▇ 0.8%	at its level
Creative Writing v3 lab	84.4% ✎	▂····▂▆··▆█▆ beats 6/14	easy · avg 70%	▇ 0.7%	small run
SimpleQA	23.6% ✎	▄▄▄▆█▆███·▂█ beats 21/32	very hard · avg 6.4%	▇ 0.6%	typical
KOR-Bench	70.2% ✎	▄·······█··▄ beats 1/3	medium · avg 55%	▇ 0.6%	at its level
NOVA-63 lab	52.5% ✎	▃▃·▃···▃█▃▃█ beats 3/11	medium · avg 38%	▇ 0.6%	blunt · at its level
RULER 128k	94.5% ✎	▃··▅▄▇▅█▇·▃▃ beats 31/32	easy · avg 72%	▇ 0.4%	typical
Chess Puzzles (Epoch)	8.0% ✎	█▁▁▃▁▂▁▄▂▂▁▁ beats 54/135	very hard · avg 5.9%	▇ 0.4%	sharp · too hard for it · noisy
BeyondAIME	56.0% ✎	▄·▂·▂▄··▄▄▄█ beats 6/15	medium · avg 36%	▇ 0.2%	at its level
DeepPlanning lab	4.9% ✎	▂·····▃··▄█▄ beats 1/9	very hard · avg 7.1%	▇ 0.2%	noisy
Vectara hallucination leaderboard, 2025 dataset (HHEM-2.3, factual consistency rate)	95.2% ✎	▁▁▂▃▄▄█▆▃▂▂▁ beats 64/140	very easy · avg 94%	▇ 0.1%	blunt · maxed out · noisy
miniF2F (pass@1)	5.7% ✎	▃·█▅·····▅·▃ beats 0/8	very hard · avg 14%	▇ 0.1%	blunt · noisy · small run
WMT24++ lab	69.3% ✎	▁·▁▁·▁·▁▃▆█▃ beats 5/27	easy · avg 68%	▇ 0.1%	blunt · at its level

$ where to run it

Through an API

api · Alibaba Cloud Model Studio · $0.75/M · $0.2 in / $2.4 out per 1M tokens · 262k context · www.alibabacloud.com ↗ · price sourceUSD per 1M tokens, Alibaba Cloud Model Studio, Singapore (International deployment scope), qwen3-30b-a3b-thinking-2507 (thinking mode only; output = chain of thought + answer). Chinese-mainland/Global deployments priced differently; free quota, batch and context-cache discounts excluded. Currency: USD. Alibaba Cloud official Model Studio pricing page, observed 2026-09-25.

llama.cpp · recommended

version

Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
Windows
irm https://llama.app/install.ps1 | iex
other ways to install ↗
Run Qwen3 30B-A3B Thinking 2507 · needs ~18.6 GB · from huggingface.co/lmstudio-community
chat in the terminal
llama cli -hf lmstudio-community/Qwen3-30B-A3B-Thinking-2507-GGUF:Q4_K_M
Or start a local server and open the address it prints (http://localhost:9931)
llama serve -hf lmstudio-community/Qwen3-30B-A3B-Thinking-2507-GGUF:Q4_K_M

LM Studio

version

Install LM Studio · Open the app once before using the lms command.
download it (macOS / Windows / Linux) ↗
Run Qwen3 30B-A3B Thinking 2507 · needs ~18.6 GB · from huggingface.co/lmstudio-community
open in LM Studio ↗
Or download it in the terminal
lms get https://huggingface.co/lmstudio-community/Qwen3-30B-A3B-Thinking-2507-GGUF@Q4_K_M -y
Then run lms ls to see its name, and lms chat followed by that name

Ollama

version

Install Ollama
download it (macOS / Windows) ↗
Linux
curl -fsSL https://ollama.com/install.sh | sh
Run Qwen3 30B-A3B Thinking 2507 · needs ~18.8 GB · from ollama.com
chat in the terminal
ollama run qwen3:30b

MLX (Mac)

version

Install MLX (Mac) · Apple silicon Macs only, text only.
with uv
uv tool install mlx-lm
or
pip install mlx-lm
Run Qwen3 30B-A3B Thinking 2507 · needs ~17.2 GB, Mac only · from huggingface.co/lmstudio-community
chat in the terminal
mlx_lm.chat --model lmstudio-community/Qwen3-30B-A3B-Thinking-2507-MLX-4bit
Or start an OpenAI-compatible server at http://localhost:8080/v1
mlx_lm.server --model lmstudio-community/Qwen3-30B-A3B-Thinking-2507-MLX-4bit

Official weights and documentation: huggingface.co/Qwen ↗

$ how the score was worked out · Qwen/Qwen3-Next-80B-A3B-Think… 23%, Qwen/Qwen3-30B-A3B-Thinking-2… 22%, ByteDance-Seed/Seed-OSS-36B-I… 18% and 11 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
Qwen/Qwen3-Next-80B-A3B-Thinking model card · lab-reported	19	▇▇▇▇▇▇▇ 23%	apache-2.0
Qwen/Qwen3-30B-A3B-Thinking-2507 model card · lab-reported	23	▇▇▇▇▇▇▇ 22%	apache-2.0
ByteDance-Seed/Seed-OSS-36B-Instruct model card	18	▇▇▇▇▇ 18%	apache-2.0
Qwen/Qwen3.5-9B model card · lab-reported	16	▇▇▇▇ 13%	apache-2.0
Epoch AI — internal evaluations	3	▇▇ 6%	CC BY 4.0
Qwen/Qwen3.5-4B model card · lab-reported	3	▇▇ 5%	apache-2.0
MathArena ArXivMath / BrokenArXiv / ArXivLean model-output datasets	3	▇ 5%	CC BY-SA 4.0
Qwen (Alibaba) vendor-reported benchmark scores for its own models (qwen.ai launch posts, Qwen2.5-VL and Qwen3-VL technical reports, Qwen3.8-Flash-Next card) · lab-reported	1	▇ 3%	Factual vendor-reported score (not licensed content)
Medmarks: A Comprehensive Open-Source LLM Benchmark Suite for Medical Tasks — authors' own runs of 61 models	1	▇ 2%	CC BY 4.0
Knowledge Index of Noah's Ark (arXiv:2606.05104v2)	1	▇ 1%	CC BY 4.0
BRIDGE Medical Leaderboard (Harvard Medical School / Brigham and Women's YLab)	1	▇ 1%	MIT
NVIDIA — Nemotron 3 Nano technical report	2	▇ 1%	CC BY 4.0
LiquidAI/LFM2.5-8B-A1B model card	1	▇ 0%	LFM Open License v1.0
Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository)	1	▇ 0%	Apache-2.0
prior · pull toward the middle		▇ 0%
weighted average		= 60.6

$ quantization evidence · 12 measurements

How much of the full-precision result a quantized version kept, measured within one study. Rows from the same study share a table and aren't independent.

precision	test	full → quantized	kept
INT4 W4A16 (weight-only INT4, symmetric, group 128; GPTQ, llm-compressor)	MATH-500 — published row 'Math 500'; lighteval (litellm endpoint to vLLM server), temperature 0.6, top_p 0.95, top_k 20, max_new_tokens 32000	0.907 → 0.931	102.6%	source ↗
INT4 W4A16 (weight-only INT4, symmetric, group 128; GPTQ, llm-compressor)	IFEval — published row 'IfEval (0-shot)'; lm-evaluation-harness task ifeval via vLLM OpenAI-compatible server (backend local-chat-completions, chat template, --num_fewshot 0, gen_kwargs do_sample=True,temperature=0.6,top_p=0.95,top_k=20,min_p=0.0,max_gen_toks=64000)	0.901 → 0.903	100.2%	source ↗
INT4 W4A16 (weight-only INT4, symmetric, group 128; GPTQ, llm-compressor)	GPQA Diamond — published row 'GPQA diamond'; lighteval (litellm endpoint to vLLM server), temperature 0.6, top_p 0.95, top_k 20, max_new_tokens 32000	0.715 → 0.716	100.1%	source ↗
INT4 W4A16 (weight-only INT4, symmetric, group 128; GPTQ, llm-compressor)	MMLU-Pro — published row 'MMLU-Pro (0-shot)'; lm-evaluation-harness task mmlu_pro_chat via vLLM OpenAI-compatible server (backend local-chat-completions, chat template, --num_fewshot 0, gen_kwargs do_sample=True,temperature=0.6,top_p=0.95,top_k=20,min_p=0.0,max_gen_toks=64000)	0.807 → 0.791	98.0%	source ↗
INT4 W4A16 (weight-only INT4, symmetric, group 128; GPTQ, llm-compressor)	AIME 2025 — published row 'AIME 2025'; lighteval (litellm endpoint to vLLM server), temperature 0.6, top_p 0.95, top_k 20, max_new_tokens 32000	0.867 → 0.833	96.1%	source ↗
INT4 W4A16 (weight-only INT4, symmetric, group 128; GPTQ, llm-compressor)	GSM8K-Platinum — published row 'GSM8k Platinum (0-shot)'; lm-evaluation-harness task gsm8k_platinum_cot_llama via vLLM OpenAI-compatible server (backend local-chat-completions, chat template, --num_fewshot 0, gen_kwargs do_sample=True,temperature=0.6,top_p=0.95,top_k=20,min_p=0.0,max_gen_toks=64000)	0.856 → 0.727	84.9%	source ↗
INT8 W8A8 (weights INT8 per-channel, activations INT8 per-token dynamic; SmoothQuant+GPTQ, llm-compressor)	GPQA Diamond — published row 'GPQA diamond'; lighteval (litellm endpoint to vLLM server), temperature 0.6, top_p 0.95, top_k 20, max_new_tokens 32000	0.715 → 0.724	101.3%	source ↗
INT8 W8A8 (weights INT8 per-channel, activations INT8 per-token dynamic; SmoothQuant+GPTQ, llm-compressor)	IFEval — published row 'IfEval (0-shot)'; lm-evaluation-harness task ifeval via vLLM OpenAI-compatible server (backend local-chat-completions, chat template, --num_fewshot 0, gen_kwargs do_sample=True,temperature=0.6,top_p=0.95,top_k=20,min_p=0.0,max_gen_toks=64000)	0.901 → 0.901	100.0%	source ↗
INT8 W8A8 (weights INT8 per-channel, activations INT8 per-token dynamic; SmoothQuant+GPTQ, llm-compressor)	MMLU-Pro — published row 'MMLU-Pro (0-shot)'; lm-evaluation-harness task mmlu_pro_chat via vLLM OpenAI-compatible server (backend local-chat-completions, chat template, --num_fewshot 0, gen_kwargs do_sample=True,temperature=0.6,top_p=0.95,top_k=20,min_p=0.0,max_gen_toks=64000)	0.807 → 0.807	99.9%	source ↗
INT8 W8A8 (weights INT8 per-channel, activations INT8 per-token dynamic; SmoothQuant+GPTQ, llm-compressor)	MATH-500 — published row 'Math 500'; lighteval (litellm endpoint to vLLM server), temperature 0.6, top_p 0.95, top_k 20, max_new_tokens 32000	0.907 → 0.906	99.9%	source ↗
INT8 W8A8 (weights INT8 per-channel, activations INT8 per-token dynamic; SmoothQuant+GPTQ, llm-compressor)	GSM8K-Platinum — published row 'GSM8k Platinum (0-shot)'; lm-evaluation-harness task gsm8k_platinum_cot_llama via vLLM OpenAI-compatible server (backend local-chat-completions, chat template, --num_fewshot 0, gen_kwargs do_sample=True,temperature=0.6,top_p=0.95,top_k=20,min_p=0.0,max_gen_toks=64000)	0.856 → 0.838	98.0%	source ↗
INT8 W8A8 (weights INT8 per-channel, activations INT8 per-token dynamic; SmoothQuant+GPTQ, llm-compressor)	AIME 2025 — published row 'AIME 2025'; lighteval (litellm endpoint to vLLM server), temperature 0.6, top_p 0.95, top_k 20, max_new_tokens 32000	0.867 → 0.842	97.1%	source ↗

$ improve this page

Update model information Add a place to run it Add a test result Propose a benchmark Correct a benchmark score Edit data on GitHub

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 2026-10-11 02:20 UTC · sources and licences · json