AlibabaProprietary

Qwen3.6-Max-Preview

Qwen3.6-Max-Preview is a large-scale foundation model developed by Alibaba. With an extremely high computational capacity of approximately 10 trillion parameters, it is a chat-focused model supporting a maximum context window of 262K tokens.

Parameters

10000.0B

Context Window

262K

License

Proprietary

Release Date

2026-04-20

API Pricing

Input Price (per 1M tokens)

$1.3

Output Price (per 1M tokens)

Billing Mode: standard

Strengths

・Massive 10-trillion scale parameter count
・Extensive 262K context window
・Advanced conversational AI capabilities

Weaknesses

・Proprietary model weights are not public
・Restrictive closed-source licensing
・Lack of detailed benchmark results

Use Cases

・Analysis of ultra-long documents
・Building complex conversational AI services
・Advanced processing of large-scale data

Deep Analysis

Artificial Analysis Intelligence Index

#2 overall out of 201 models as of April 2026

SWE-Bench Pro

57.3%

Top score on this benchmark, per Alibaba claims

Terminal-Bench 2.0

65.4%

Ties with Claude Opus 4.6

Input Price

$6/1M tokens

Standard pricing via Alibaba Cloud; free during preview period

Context Window

262K tokens

Large, but below 1M of some competitors like Qwen3.6-Plus

Total Parameters

>1 trillion

MoE architecture; active parameters not disclosed

Strengths

・Exceptional agentic coding performance, topping six major benchmarks including SWE-bench Pro and Terminal-Bench 2.0.
・Competitive pricing relative to Western frontier models, with significant cost savings.
・Advanced features like preserve_thinking for reliable multi-step agent workflows and improved instruction following.

Weaknesses

・Closed weights and API-only access, preventing self-hosting and limiting transparency.
・Preview model with no production SLA, subject to changes in behavior and pricing.
・Text-only modality at launch, lacking image input support compared to some competitors.

Competitor Comparison

Model	Arena	SWE	GPQA	Price
GPT-5.4	89 (BenchLM composite)	57.7% (SWE-Bench Pro)	N/A	$2.50/$14.00
Claude Opus 4.6	N/A	80.8% (SWE-Bench Verified)	~90%	$15.00/$75.00
Kimi K2.6	N/A	80.2% (SWE-Bench Verified)	N/A	$0.60/$3.00

Sources

Analysis generated: 2026-05-23