Qwen3-Coder-Next 80B-A3B (MoE)
Qwen3-Coder-Next 80B-A3B (MoE) needs 48.9 GB of VRAM at Q4_K_M with 8k context, computed from its config.json as checked on 3 November 2026. The smallest card that holds it has 80 GB, and the fastest desktop card that runs it, the RTX PRO 6000 Blackwell, generates at most 721 tokens per second.
- Parameters
- 79.7B, 3B active
- Context
- 262,144 tokens
- Licence
- Apache 2.0
- Released
- January 2026
- Config checked
- 3 November 2026
Open Qwen3-Coder-Next 80B-A3B (MoE) in the LLM checker
Which card size runs it
48.9 GB against each memory size. A model may use 95% of a card, and only what is left over, or missing, is coloured; tight means above 85%, with little room for a longer conversation.
Other quantisations, context lengths or a Q8 KV cache: the VRAM calculator.
Computed 3 November 2026Which GPU to pick
The smallest card here that loads it at Q4_K_M with 8k of context is the A100 (80 GB): 48.9 GB needed, an estimated 820 tokens per second. The same card still has room at 32,768 tokens of context (49.5 GB), so there is no reason to buy above it for this model.
At Q8_0, which is near-lossless, the card is the RTX PRO 6000 Blackwell (96 GB), needing 88.6 GB. On Apple silicon, a MacBook Pro M4 Max holds it with 32,768 tokens of context at an estimated 220 tokens per second; prompt processing is slower than on Nvidia, which long documents make obvious. Among unified-memory desktops, the Ryzen AI Max+ 395 holds it at 32,768 tokens in 49.5 GB, at an estimated 103 tokens per second.
Speed on each card
| GPU | Fits | Est. tokens/s | Feels like | Memory | Bandwidth |
|---|---|---|---|---|---|
| H100 SXM | Yes | ≤ 1,348 | faster than you can read | 80 GB | 3,350 GB/s |
| A100 | Yes | ≤ 820 | faster than you can read | 80 GB | 2,039 GB/s |
| RTX PRO 6000 Blackwell | Yes | ≤ 721 | faster than you can read | 96 GB | 1,792 GB/s |
| RTX Spark | Yes | ≤ 121 | faster than you can read | 128 GB (102 for the GPU) | 300 GB/s |
| DGX Spark | Yes | ≤ 110 | faster than you can read | 128 GB (126 for the GPU) | 273 GB/s |
| Ryzen AI Max+ 395 | Yes | ≤ 103 | faster than you can read | 128 GB (96 for the GPU) | 256 GB/s |
| RTX 6000 Ada | No | - | does not fit | 48 GB | 960 GB/s |
| L40S | No | - | does not fit | 48 GB | 864 GB/s |
| RTX A6000 | No | - | does not fit | 48 GB | 768 GB/s |
| RTX PRO 5000 Blackwell | No | - | does not fit | 48 GB | 1,344 GB/s |
| RTX 5090 | No | - | does not fit | 32 GB | 1,792 GB/s |
| RTX PRO 4500 Blackwell | No | - | does not fit | 32 GB | 896 GB/s |
| Radeon AI PRO R9700 | No | - | does not fit | 32 GB | 640 GB/s |
| Arc Pro B70 | No | - | does not fit | 32 GB | 608 GB/s |
| RTX 3090 | No | - | does not fit | 24 GB | 936 GB/s |
| RTX 4090 | No | - | does not fit | 24 GB | 1,008 GB/s |
| RX 7900 XTX | No | - | does not fit | 24 GB | 960 GB/s |
| RTX PRO 4000 Blackwell | No | - | does not fit | 24 GB | 672 GB/s |
| RTX PRO 4000 Blackwell SFF | No | - | does not fit | 24 GB | 432 GB/s |
| Arc Pro B60 | No | - | does not fit | 24 GB | 456 GB/s |
| RTX 5090 Laptop GPU | No | - | does not fit | 24 GB | 896 GB/s |
| RX 7900 XT | No | - | does not fit | 20 GB | 800 GB/s |
| RTX 4060 Ti | No | - | does not fit | 16 GB | 288 GB/s |
| RTX 5060 Ti | No | - | does not fit | 16 GB | 448 GB/s |
| RTX 5080 | No | - | does not fit | 16 GB | 960 GB/s |
| RTX 5070 Ti | No | - | does not fit | 16 GB | 896 GB/s |
| RTX 4080 SUPER | No | - | does not fit | 16 GB | 736 GB/s |
| RX 9070 XT | No | - | does not fit | 16 GB | 640 GB/s |
| RX 9060 XT | No | - | does not fit | 16 GB | 320 GB/s |
| RTX 3060 | No | - | does not fit | 12 GB | 360 GB/s |
| RTX 4070 | No | - | does not fit | 12 GB | 504 GB/s |
| RTX 5070 | No | - | does not fit | 12 GB | 672 GB/s |
| Arc B580 | No | - | does not fit | 12 GB | 456 GB/s |
Single-stream decode, from memory bandwidth at 70% efficiency. Prompt processing and batching not included. These are ceilings, not measurements: Qwen3-Coder-Next 80B-A3B (MoE) reads only 1.7 GB of weights per token at Q4, so little that the per-token costs the formula leaves out decide much of the real speed. The 6 figures marked ≤ are where a real runtime falls furthest below the number shown. See the speed estimator for other quantisations and Apple hardware, or every card compared if you are choosing hardware rather than a model.
Computed 3 November 2026Memory by quantisation and context
| Quant | 4,096 ctx | 8,192 ctx | 32,768 ctx |
|---|---|---|---|
| Q8_0 | 88.5 GB · fits 96 GB | 88.6 GB · fits 96 GB | 89.2 GB · fits 96 GB |
| Q5_K_M | 59.5 GB · fits 80 GB | 59.6 GB · fits 80 GB | 60.2 GB · fits 80 GB |
| Q4_K_M | 48.8 GB · fits 80 GB | 48.9 GB · fits 80 GB | 49.5 GB · fits 80 GB |
Weights at this quant: 46.2 GB at Q4. Every extra 1,000 tokens of context adds about 0.025 GB of KV cache at FP16, on top of a fixed 0.08 GB of recurrent state that does not grow. Try other settings in the VRAM calculator.
About the model
Qwen3-Coder-Next 80B-A3B (MoE) has 79.7B parameters with 3B active per token (mixture of experts). It has 48 layers, of which only 12 are full attention (2 KV heads of dimension 256) and keep a cache that grows with the conversation; the other 36 are Gated DeltaNet linear-attention layers with a small fixed state. Alibaba's coding-agent model, early 2026: 80B of weights across 512 experts, with 3B read per token. Built for coding agents and local development, and designed to plug into CLI and IDE tools.
Text only 262,144-token window Apache 2.0 January 2026 3B of 79.7B active
Reasoning. Non-thinking only: it answers directly and never emits reasoning blocks.
All 512 experts stay in memory; 3B are active per token, so it answers like a small model once it is loaded.
This is a coding model, tuned for writing and editing code and for agentic tools that read whole repositories, which is where a long context earns its memory.
Native context window: 262,144 tokens. Licence: Apache 2.0, a permissive licence. Weights published in January 2026. It follows Qwen2.5 Coder 32B.
Good at
- A coding agent that answers at small-model speed, on an 80 or 96 GB card or a large-memory Mac.
- Repository-scale context: a 262,144-token window with 12 cache-growing layers out of 48.
- Agent loops, where no thinking preamble means no wasted tokens per step.
Watch out for
- It is the memory that costs, not the compute. It does not fit a 48 GB card at Q4.
- A specialist. It is not the model for general conversation.
- The hybrid layers need a current runtime.
Similar models
The models that need about the same memory as Qwen3-Coder-Next 80B-A3B (MoE) at Q4_K_M and 8k, which makes them the real alternatives on whatever card you have. The context column is what each extra 1,000 tokens costs, and it is where models of the same size differ most.
| Model | Size | Needs | Per 1k context | Window | Released |
|---|---|---|---|---|---|
| Qwen3-Coder-Next 80B-A3B (MoE) | 79.7B · 3B active | 48.9 GB | 0.025 GB | 256k | 2026-01 |
| Nemotron 3.5 Bolt 74B-A9B (MoE) | 74.9B · 9B active | 45.9 GB | 0.008 GB | 256k | 2026-10 |
| DeepSeek-R1 Distill Llama 70B | 70.6B | 45.8 GB | 0.328 GB | 128k | 2025-01 |
| Llama 3.3 70B | 70.6B | 45.8 GB | 0.328 GB | 128k | 2024-12 |
| gpt-oss 120B | 117B · 5.1B active | 71.4 GB | 0.037 GB | 128k | 2025-08 |
See every model compared, which models fit on 8 to 96 GB, or the cheapest hour of every cloud card.
Questions
- Can I run Qwen3-Coder-Next 80B-A3B (MoE) on a 24 GB card?
- Not comfortably. At Q4_K_M and 8k context it needs about 48.9 GB. The smallest tier that fits is 80 GB.
- How much VRAM does Qwen3-Coder-Next 80B-A3B (MoE) need at Q8?
- About 88.6 GB at 8k context, or 89.2 GB at 32k. Q8 is near-lossless; use it when it fits.
- What is the context length of Qwen3-Coder-Next 80B-A3B (MoE)?
- 262,144 tokens natively. Holding all of it at Q4_K_M takes about 55.1 GB, of which 6.5 GB is context.
- Why does Qwen3-Coder-Next 80B-A3B (MoE) need so little memory for long context?
- Because most of its layers do not keep a cache that grows. It has 48 layers, of which only 12 are full attention (2 KV heads of dimension 256) and keep a cache that grows with the conversation; the other 36 are Gated DeltaNet linear-attention layers with a small fixed state. At 32k tokens the context costs about 0.9 GB; if all 48 layers kept a conventional full-length cache, the same conversation would cost about 3.2 GB.
- How fast is Qwen3-Coder-Next 80B-A3B (MoE) on an RTX 4090?
- It does not fit a 4090 at Q4 and 8k context, so speed would collapse to CPU offloading. Use a larger card or a smaller quantisation.
- Architecture values from config.json in Qwen/Qwen3-Coder-Next on Hugging Face, and its model card, checked 3 November 2026. Every figure on this page is computed from those values; check them before buying hardware for this model.
- Estimates assume a single conversation on a card that is otherwise free. A desktop on the same GPU takes 0.5 to 2 GB.