Qwen/Qwen3.5-35B-A3B-FP8
Parameters36B (3B active)
Context256K
Input → outputimage + text → text
Licenseapache-2.0
PrecisionF8_E4M3 · fp8
ArchitectureQwen3_5MoeForConditionalGeneration
ReleasedFeb 25, 2026
Downloads1.3M
Based on Qwen/Qwen3.5-35B-A3B
Other sizes and formats in the family
| Model | Params | Precision | Downloads |
|---|---|---|---|
| Qwen/Qwen3.5-397B-A17B-GPTQ-Int4gptq | 397B | – | 2.7K |
| Qwen/Qwen3.5-122B-A10B-FP8fp8 | 125B | F8_E4M3 | 436.9K |
| Qwen/Qwen3.5-122B-A10B | 125B | BF16 | 699.1K |
| Qwen/Qwen3.5-122B-A10B-GPTQ-Int4gptq | 125B | I32 | 368.1K |
| Qwen/Qwen3.5-35B-A3B-GPTQ-Int4gptq | 36B | I32 | 260.5K |
| Qwen/Qwen3.5-27B-FP8fp8 | 27.8B | F8_E4M3 | 149.7K |
| Qwen/Qwen3.5-27B-GPTQ-Int4gptq | 27.8B | I32 | 17.5K |
| Qwen/Qwen3.5-9B | 9.7B | BF16 | 7.8M |
| Qwen/Qwen3.5-9B-Base | 9.7B | BF16 | 373.7K |
| Qwen/Qwen3.5-4B | 4.7B | BF16 | 7.9M |
| Qwen/Qwen3.5-4B-Base | 4.7B | BF16 | 739.3K |
| Qwen/Qwen3.5-2B | 2.3B | BF16 | 2.8M |
| Qwen/Qwen3.5-2B-Base | 2.3B | BF16 | 315.3K |
| Qwen/Qwen3.5-0.8B | 873M | BF16 | 3.1M |
| Qwen/Qwen3.5-0.8B-Base | 873M | BF16 | 484K |
This release in the news
No coverage of this exact release yet.
Recent news about Qwen
- Oct 19, 2026 · Opinion / analysisA bare IPython harness ties production coding agents with Qwen3.5-9B on TB-Lite, and tool-output normalization fixed most of the run-to-run variance
- Oct 19, 2026 · Opinion / analysisKapitsa: open-source Apple Silicon optimized Metal engine built for Qwen and Nemotron. See benchmarks of Qwen 27B, Nemotron + 8 other models on an M1 Pro (16GB), M4 Pro (48GB) and M5 Max (128GB)
- Oct 19, 2026 · Opinion / analysisBasic Workflow Simple Prompt Testing: Krea2 vs MOSS vs Qwen2.1 vs Flux3
- Oct 19, 2026 · Opinion / analysisCalling all GPUs: help us benchmark SonicBoom v0.0.3 — a compiler-first AI runtime that beats llama.cpp on decode
- Oct 19, 2026 · Model releaseNemotron 3.5 Bolt, Muse Spark Open, and a Fast LFM2.5 24B
- Oct 19, 2026 · Opinion / analysisMOSS Character LoRas
- Oct 19, 2026 · Opinion / analysisA comparison of 6 open source models. Krea2, Qwen2.1, Flux3 Klein, Flux 2 Klein, ZIT, & Moss. Same prompt used for all.
- Oct 19, 2026 · Benchmark resultTogether Post-Training: frontier-grade model shaping, the way that fits you