The Open-Source AI Stack
RSS
All models

Models · qwen3

Qwen3 Max

Proprietary Alibaba · — · Proprietary

Trillion-parameter MoE, API-only via Qwen Chat and Alibaba Cloud at release. Pretrained on roughly 36T tokens. Alibaba's first proprietary Qwen flagship at this scale, breaking the open-weights pattern. Supports more than 100 languages.

Cost

$1.66 / Mtok input
$7.22 / Mtok output

· as of 2026-05-21

source ↗

Speed

32.4 tok/sec output
1883 ms TTFT

· as of 2026-05-21

source ↗

Architecture

tokens in Embedding vocab not disclosed × N layers Attention (not disclosed) Position encoding not disclosed context not disclosed MoE Router ? experts total · ? active per token Output projection tokens out
Schema-generated from data/models.yaml. Every label is auditable against the model's sources.

Specs

Architecture
moe
Total params
not disclosed
Active params
not disclosed
Context window
not verified
Attention
Position encoding
Pretraining tokens
36.0T
Post-training
sft, rlhf
OSI-approved
no
Data released
no
Training code
not released

Benchmarks

Each score carries the date it was published; we never infer or interpolate missing scores.

General reasoning

MMLU-Pro 84.1 as of 2026-05-21 source ↗
GPQA-Diamond 76.4 as of 2026-05-21 source ↗

Available quantizations

None. The weights are not distributed, so there are no public quantizations.

Notable innovations

  • · First trillion-parameter Qwen
  • · 36T-token pretraining run
  • · Proprietary tier of the Qwen line

Lineage

Proprietary trillion-parameter Qwen variant.

Derived from

Qwen 3 235B A22B Instruct 2025-04-28

Sources