Alibaba · Open weightsQwen: Qwen3.8 Omni Flash

Active

Model portfolio

Qwen: Qwen3.8 Omni Flash

Tap a star to rate

Alibaba · Released 2026-09-21 · v2026-09-21 · Open · Context 1M tokens

Qwen: Qwen3.8 Omni Flash is an open-source model from Alibaba (released 2026-09-21). Context window ~1M tokens with capabilities: Vision, Audio, Video, Text. List pricing ≈ $0.15 / $0.47 per 1M input/output tokens. Top tracked signals: MATH · benchlm 96.2, MATH 96.2, LIVECODEBENCH · benchlm 92.6. Cadle merges provider listings, unified leaderboard medians, and public pricing where available.

ACTIVEVisionAudioVideoText

Language Cap

Cadle Language #33/385 · 52%

Benchmark average

86.8

#19 on the benchmark matrix

Pricing

$0.15

$0.47 / 1M out

Compare model ProviderDocs ↗

CROSS-BOARD RANKS

Cadle Language LLM #33/385Matrix #19/443

CADLE CAP vs MODEL MATRIX

Cadle Cap ranks models on every published score we have for them — harder benchmarks weigh more. Missing tests are skipped, not zeroed. It answers who is stronger than whom in today's pool.

Model Matrix is the datasheet: Overall, Intelligence, Coding, Agentic, MMLU, GPQA, Math, LiveCodeBench, HumanEval, Reason, ARC-AGI. A dash means unmeasured. AVG is an importance-weighted mean of that model's filled cells only.

MODEL MATRIX · TRUTH CENTER CELLS

Overall

—

AA IQ

—

Coding

80.5

Agentic

87.1

MMLU

59.4

GPQA

91.0

Math

96.2

LiveCodeBench

92.6

HumanEval

—

Reason

91.6

ARC-AGI

—

Coverage AVG

86.8

Cadle Cap ranks models with within-benchmark percentiles so ARC-AGI ~40 and MMLU ~90 never compete raw. Coverage AVG is an importance-weighted mean of filled expert cells only — blanks are skipped, not zeroed.

ANALYSIS · TRUTH CENTER

Hard benches (GPQA · ARC-AGI · LCB · Agentic) sit left — a mid-40 on ARC-AGI can still be frontier. Broader benches on the right are easier and often higher.

Full Matrix desk →

CADLE INSIGHTS

Only measured signals

CADLE CAP

Cadle #33/385 · 52% confidence

CODING

80.5

Media Cap quality axis

AGENTIC

87.1

Media Cap quality axis

MATH

96.2

Media Cap quality axis

KNOWLEDGE

80.2

Media Cap quality axis

BENCHMARK AVG

85.5

14 tracked scores

STRONGEST SIGNAL

MATH · benchlm

Score 96.2

BLENDED $/1M

$0.62

In $0.15 · Out $0.47

SPEED · COST · QUALITY

Measured matrix / AA / gateway rows only.

Cap Coding: 93.7
Matrix Coding: 80.5
Input $/1M: $0.15
Output $/1M: $0.47

DESCRIPTION + DETAILS

PROVIDER

Alibaba

RELEASE

2026-09-21

CONTEXT

1M tokens

STATUS

ACTIVE

COMMUNITY REVIEWS

OPERATOR NOTES

Quality, latency, cost, and failure modes — not marketing copy.

No reviews yet

WRITE A REVIEW

Tap a star to rate

Stored on this device · synced when signed in

No operator notes yet

First review sets the tone for how teams evaluate Qwen: Qwen3.8 Omni Flash.

MODEL COMPARISON

Metric	Qwen: Qwen3.8 Omni Flash
Provider	Alibaba
Release	2026-09-21
Input / 1M	$0.15
Output / 1M	$0.47
Context	1M tokens
Cadle Cap	91
Cap board	Cadle Language #33/385
MATH · benchlm	96.2
MATH	96.2
LIVECODEBENCH · benchlm	92.6
LIVECODEBENCH	92.6
REASON · benchlm	91.6
REASON	91.6
GPQA · benchlm	91.0
GPQA	91.0
AGENTIC · benchlm	87.1
AGENTIC	87.1
CODING · benchlm	80.5
CODING	80.5