AION
Modelofficial release

Qwen/Qwen3.5-4B-Base

Qwen →Hugging Face ↗
Parameters4.7B
Context256K
Input → outputimage + text → text
Licenseapache-2.0
PrecisionBF16
ArchitectureQwen3_5ForConditionalGeneration
ReleasedFeb 27, 2026
Downloads653.9K

Other sizes and formats in the family

ModelParamsPrecisionDownloads
Qwen/Qwen3.5-397B-A17B-GPTQ-Int4gptq397B–3K
Qwen/Qwen3.5-122B-A10B-FP8fp8125BF8_E4M3507.5K
Qwen/Qwen3.5-122B-A10B125BBF16594.2K
Qwen/Qwen3.5-122B-A10B-GPTQ-Int4gptq125BI32336.1K
Qwen/Qwen3.5-35B-A3B-FP8fp836BF8_E4M31.3M
Qwen/Qwen3.5-35B-A3B-GPTQ-Int4gptq36BI32203.5K
Qwen/Qwen3.5-27B-FP8fp827.8BF8_E4M3168.8K
Qwen/Qwen3.5-27B-GPTQ-Int4gptq27.8BI3217.8K
Qwen/Qwen3.5-9B9.7BBF168.3M
Qwen/Qwen3.5-9B-Base9.7BBF16304.1K
Qwen/Qwen3.5-4B4.7BBF167.9M
Qwen/Qwen3.5-2B2.3BBF163.2M
Qwen/Qwen3.5-2B-Base2.3BBF16305.8K
Qwen/Qwen3.5-0.8B873MBF162.6M
Qwen/Qwen3.5-0.8B-Base873MBF16479.7K

Compare side by side →

Built on this model

ModelParamsPrecisionDownloads
Qwen/Qwen3.5-4B4.7BBF167.9M

This release in the news

No coverage of this exact release yet.

Recent news about Qwen

  1. Oct 11, 2026 · Opinion / analysis
    Reverse Engineering w/ Local?
  2. Oct 11, 2026 · Opinion / analysis
    UPDATE: Qwen 3.8 27B 140 tok/s on single RTX 3090 Megakernel: KL divergence 0.0009 vs llama.cpp
  3. Oct 11, 2026 · Model release
    Qwen/Qwen-Image-2.1-Turbo
  4. Oct 11, 2026 · Opinion / analysis
    Building prompts with LLMs
  5. Oct 11, 2026 · Opinion / analysis
    [P] Pecision models that score every allowed label from the logits: Jebadiah v2.1 (27B, 9B), open weights and self-run benchmark results [P]
  6. Oct 11, 2026 · Opinion / analysis
    PSA: DeepSeek V4.1 Flash habitually exfiltrates API keys. It is dangerously misaligned and may be hazardous to use
  7. Oct 11, 2026 · Opinion / analysis
    Qwen 3.8 27B Q5 vs Qwen 3.8 Next Q3_S for document analysis
  8. Oct 11, 2026 · Opinion / analysis
    Heretic or Abliterated?