Model Marketplace

Browse available models, pricing, and model names. Price unit: USD/1M tokens

All25 DeepSeek2 OpenAI9 Moonshot5 Alibaba5 Anthropic4

Alibaba updates

News · Top ranked

arxiv.orgSep 3, 2026, 07:30 AM

Post-Training Ternarization Compresses Qwen3-4B but Reduces Capability

An arXiv study finds that post-training ternarization can reduce Qwen3-4B storage substantially, but capability and perplexity worsen, while faster inference remains unproven.

news.aibase.comAug 27, 2026, 08:02 AM

Alibaba Qwen Open-Sources Qwen3.8-Flash-Next Architecture

Alibaba’s Qwen team has released the Qwen3.8-Flash multimodal MoE model and open-sourced the Qwen3.8-Flash-Next architecture, an early foundation for Qwen4. The 125B-parameter model activates 6B parameters per token, supports 260,000 tokens natively, and can extend to 1 million tokens with YaRN. Qwen says training costs are one-ninth those of Qwen3.7-Plus. Its architecture combines GDN and QSA attention, gated residual streams, N-gram Embedding, and mixed Muon-AdamW optimization. Weights and a technical report are available on Hugging Face and ModelScope.

Updated Jul 10, 2026, 08:17 AM

20%

Qwen3.6 Max Preview

qwen3.6-max-preview

Standard

Input$1.9

Output$11.4

Cache read$0

Cache write$0

Batch input$1

Batch output$1

Capabilities

chat/completions

Provider: Alibaba
Max output: 1,000,000 tokens
Protocol: OpenAI-compatible

Updated Jul 10, 2026, 08:17 AM

20%

Qwen3.7 Max

qwen3.7-max

Standard

Input$1.58

Output$4.74

Cache read$0

Cache write$0

Batch input$1

Batch output$1

Capabilities

chat/completionsresponses

Provider: Alibaba
Max output: 1,000,000 tokens
Protocol: OpenAI-compatible

Updated Jul 10, 2026, 08:17 AM

20%

Qwen3.7 Plus

qwen3.7-plus

Standard

Input$0.79

Output$3.16

Cache read$0

Cache write$0

Batch input$1

Batch output$1

Capabilities

chat/completionsresponses

Provider: Alibaba
Max output: 1,000,000 tokens
Protocol: OpenAI-compatible

Updated Aug 31, 2026, 06:59 AM

Qwen3.8 Max

qwen3.8-max

Standard

Input$1.4

Output$4.2

Cache read$0.175

Cache write$1.75

Capabilities

chat/completionsresponses

Provider: Alibaba
Max output: 1,000,000 tokens
Protocol: OpenAI-compatible

Updated Aug 31, 2026, 07:03 AM

60%

Qwen Flash

qwen-flash

Standard

Input$0.03

Output$0.24

Cache read$0.006

Cache write$0

Capabilities

chat/completionsresponses

Provider: Alibaba
Max output: 1,000,000 tokens
Protocol: OpenAI-compatible