All models › alibaba
Qwen3.8-Flash-Next
alibaba · qwen-community-1.0 license · on Arena since 16 Oct 2026
Lab chat Download
Arena score1629#20 of 145 in Coding, paid models included
Size180 B61.6 B active per token (mixture of experts)
4-bit download109 GBweights only, Q4_K_M
Max context256Ktokens, from the model config
Where to use it for free
- The lab's own chat: chat.qwen.ai (free with daily limits)
- Download and run it at home: official weights · GGUF files for Ollama and LM Studio
Arena scores
| Arena | Score | Rank | Votes |
|---|---|---|---|
| Coding | 1629 | #20 of 145 | 8K |
Rank among all models, paid ones included. Source: Arena leaderboard dataset.
Speed on popular hardware
| Hardware | GPU memory | Fits? | Speed |
|---|---|---|---|
| RTX 3060 12 GB | 12 GB | Does not fit | — |
| RTX 4060 Ti 16 GB | 16 GB | Does not fit | — |
| RTX 3090 | 24 GB | Does not fit | — |
| RTX 4090 | 24 GB | Does not fit | — |
| RTX 5090 | 32 GB | Does not fit | — |
| RX 7900 XTX | 24 GB | Does not fit | — |
| 2 × RTX 3090 | 48 GB | Does not fit | — |
| 4 × RTX 3090 | 96 GB | GPU + 64 GB RAM | 7.5 tok/s |
| Mac M4 Max 128 GB | 96 GB | Does not fit | — |
| Ryzen AI Max+ 395 128 GB | 96 GB | Does not fit | — |
| 8 × H100 SXM | 640 GB | Fits in GPU | 51 tok/s |
Estimate at 4-bit with 32K context, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 70% ÷ data read per token; real numbers vary by about ±30%.
GPU memory needed
| Level | 8K context | 32K context | 128K context | Quality vs full |
|---|---|---|---|---|
| Q3_K_M | 91 GB | 91 GB | 94 GB | +4.7% noticeable loss |
| Q4_K_M | 110 GB | 110 GB | 113 GB | +1.2% very small loss |
| Q5_K_M | 129 GB | 129 GB | 132 GB | +0.4% near-lossless |
| Q6_K | 148 GB | 149 GB | 151 GB | +0.2% near-lossless |
| Q8_0 | 192 GB | 193 GB | 195 GB | ≈ 0% same as full |
| F16 | 361 GB | 361 GB | 364 GB | ≈ 0% same as full |
Weights + context cache (from the model's architecture) + runtime. Quality: increase in perplexity against the full model, from llama.cpp tests, scaled by model size. Simulate your own setup.
Coding score over time
160017001800
MayJunJulAugSeptOct
Qwen3.8-Flash-Next 1629Best paid model
Arena score from 20 Apr 2026 to 18 Oct 2026.
Similar free models in Coding
- Hy4 preview 1631
- MiMo-V2.6-Pro 1626 · 1024.2 B
- MiMo-V2.6-Flash 1633 · 310.8 B
- GLM-5.3 (Max) 1622 · 753.3 B
- DeepSeek-V4.1-Flash-20260910 (Max) 1619
- GLM-5.3-Flash 1607 · 321.3 B