What is the IQ of Qwen3.8 27B?

Qwen3.8 27B has an estimated IQ of 112, which ranks #52 out of 131 public models on the AI IQ leaderboard. The score combines six equally weighted reasoning dimensions, each measured from source-backed benchmark results.

How much does Qwen3.8 27B cost?

Qwen3.8 27B costs $0.425 per million input tokens and $3.10 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $19.03 per million input-plus-output tokens.

Is Qwen3.8 27B open source?

Yes — Qwen3.8 27B is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.

When was Qwen3.8 27B released?

Qwen3.8 27B was publicly released on August 14, 2026.

How large is the context window of Qwen3.8 27B?

Qwen3.8 27B supports a context window of 262,144 tokens — the amount of text it can consider in a single request.

Alibaba

Qwen3.8 27B

Public AI IQ model profile with source-backed benchmark results and derived ranking context.

IQ

112

EQ

n/a

Effective Cost

$19.03

IQ Rank

#52

Qwen3.8 27B API pricing — full price table, workload cost examples, and cheaper alternatives.

IQ Dimensions

Dimension	IQ	Coverage
Abstract Reasoning	87	0/3
Mathematical Reasoning	114	0/5
Academic Reasoning	132	5/6
Programmatic Reasoning	117	2/6
Computer Use	115	0/6
Reliability	107	2/7

Benchmark Results

Benchmark	Score	Dimension
AA Long Context Reasoning	77.333	Reliability
AA Omniscience	-9.98	Reliability
CritPt	5.429	Academic Reasoning
GPQA Diamond	90.505	Academic Reasoning
Humanity's Last Exam	33.92	Academic Reasoning
MMMU-Pro	76.301	Academic Reasoning
ProgramBench	1	Programmatic Reasoning
SciCode	44.676	Academic Reasoning
Terminal-Bench 2.1	79.775	Programmatic Reasoning

Compare

Frequently Asked Questions

What is the IQ of Qwen3.8 27B?: Qwen3.8 27B has an estimated IQ of 112, which ranks #52 out of 131 public models on the AI IQ leaderboard. The score combines six equally weighted reasoning dimensions, each measured from source-backed benchmark results.
How much does Qwen3.8 27B cost?: Qwen3.8 27B costs $0.425 per million input tokens and $3.10 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $19.03 per million input-plus-output tokens.
Is Qwen3.8 27B open source?: Yes — Qwen3.8 27B is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
When was Qwen3.8 27B released?: Qwen3.8 27B was publicly released on August 14, 2026.
How large is the context window of Qwen3.8 27B?: Qwen3.8 27B supports a context window of 262,144 tokens — the amount of text it can consider in a single request.