whatmodel.sh / Qwen3.8 Max (0902)

Qwen3.8 Max (0902)

+ contribute

Qwen3.8 Max (0902) is a closed-weights language model from Alibaba, released September 2026.

Catalogue identity notes

Comparison-only. Dated 0902 snapshot, kept distinct from qwen3-8-2-4t-a95b (undated 'qwen3.8-max'). Epoch model_metadata.csv: group 'Qwen3.8 Max (0902)', Alibaba, 2026-09-01, 'API access'. Arena entry 'qwen3.8-max-0902' (alibaba, Proprietary). No open-weight 0902 checkpoint verified.

76.0±0.8#34 of 719 LLMsfrom 10 benchmark results across 4 sources

$ how it compares

The LLMs ranked either side of Qwen3.8 Max (0902), out of 828 with a score, then more from Alibaba.

#	model	score	where	RAM	released
32	Gemini 3.7 Flash Google	76±1	api $1.50/M		2026-08
33	GLM-5.3 Z.ai	76±1	api $2.15/M +local	~239GB	2026-08
34	Qwen3.8 Max (0902) Alibaba	76±1	api $3.00/M		2026-09
35	DeepSeek V3.2 Speciale DeepSeek	76±3	comparison only
36	DeepSeek V4 Flash Vision (Exp) DeepSeek	76±2	local	~104GB	2026-08
	more from Alibaba
18	Qwen3.8 Max (2.4T-A95B) Alibaba	78±1	api $3.00/M +local	~380GB	2026-08
37	Qwen3.8-Flash-Next Alibaba	75±1	local	~59GB	2026-08
38	Qwen3.8-Omni-Flash Alibaba	75±2	api $0.23/M		2026-09

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #22 of 300 * · 10% of its score · 2 tests

Beats GPT-5.5, GPT-5.6 Terra, GPT-5.4 here, though they rank higher overall. Trails Grok 4.7 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Terminal-Bench 4.0	27.0% ✎	█▅▅▅▅·▅▃▃▅▅▅ beats 10/20	very hard · avg 0.4%	▇▇▇ 6.8%	sharp
Video2World V2WScore (coding agents reconstruct interactive simulated worlds from videos, OpenCode scaffold)	26.6% ✎	█··▄█·▄····▄ beats 1/6	very hard · avg 6.6%	▇ 2.9%	at its level

coding▁▁▁▃▆██▇▆▄▃▁ #11 of 458 * · 11% of its score · 1 test

Beats GPT-5.6 Sol, Claude Fable 5, GPT-5.5 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Arena WebDev (Code Arena)	1670.0 ✎	▁▂▄▃▅█▄▇▆▄▁▁ beats 110/119	avg 981	▇▇▇▇ 11%	sharp

math▁▂▃▅▆▅▄█▄▁▁▁ #25 of 398 · 56% of its score · 3 tests

Beats DeepSeek V4.1 Flash, Gemini 3.8 Flash, Grok 4.6 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
FrontierMath tiers 1–3 v2	65.6% ✎	▂··▁▁▃▄▄█▅▃▃ beats 51/82	very hard · avg 1.2%	▇▇▇▇▇▇▇▇▇▇▇ 28%	sharp · at its level
FrontierMath tier 4 v2	34.1% ✎	▁▂▁▂█▅▂▂▁▁▁▁ beats 39/63	very hard · avg 0.1%	▇▇▇▇▇▇▇▇▇▇▇ 27%	sharp · at its level
OTIS mock AIME 2024–2025	100.0% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 188/193	very hard · avg 8.5%	▇ 1.0%	sharp · maxed out · noisy

other tests4 not in a task · 23% of its score

test	result	vs the others in its run	difficulty	counts	why
Furniture Assembly (Epoch)	20.0% ✎	█▆▆▆▆▂▅▂▂▂·▅ beats 0/30	very hard · avg 1.2%	▇▇▇▇ 11%	sharp · at its level
Chess Puzzles (Epoch)	40.0% ✎	█▁▁▁▁▂▁▄▂▃▁▁ beats 118/135	very hard · avg 5.9%	▇▇ 5.5%	sharp · at its level · noisy
GPQA Diamond	92.3% ✎	▁▁▁▁▂▄▆▄▅▇█▅ beats 188/208	hard · avg 40%	▇▇ 4.4%	sharp
SimpleQA Verified	47.3% ✎	▁▁▃▂▃▃▆▆█▃▂▅ beats 48/78	very hard · avg 8.0%	▇ 2.5%	at its level

$ where to run it

Through an API

api · Alibaba Cloud Model Studio · $3.00/M · $2 in / $6 out per 1M tokens · 1M context · www.alibabacloud.com ↗ · price sourceUSD per 1M tokens, Singapore (International), dated snapshot qwen3.8-max-0902, thinking and non-thinking modes, 0<Token<=1M (single tier): input $2, output $6. Context-cache discount and free quota excluded; China (Beijing)/Global deployments ($1.65/$4.951) not used, following the International convention of other Alibaba rows. Observed 2026-09-25.

$ how the score was worked out · Epoch AI — internal evaluatio… 79%, LMArena leaderboard dataset 11%, Terminal-Bench 4.0 official l… 7% and 1 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
Epoch AI — internal evaluations	7	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 79%	CC BY 4.0
LMArena leaderboard dataset	1	▇▇▇ 11%	CC BY 4.0
Terminal-Bench 4.0 official leaderboard submissions	1	▇▇ 7%	Apache-2.0
Video2World: Benchmarking Coding Agents for Interactive World Modeling from Embodied Videos (arXiv:2610.04432v2)	1	▇ 3%	CC BY-SA 4.0
prior · pull toward the middle		▇ 0%
weighted average		= 76.0