Qwen
Qwen3 Coder 30B
A coding model for agents and editors, with tool calls turned on.
Recommended: RTX 5090, 32 GB GDDR7, $0.69 / hour, billed per minute. Check availability.
Recommended: DGX Spark, 128 GB unified, $0.79 / hour, billed per minute. Check availability.
Recommended: DGX Spark, 128 GB unified, $0.79 / hour, billed per minute. Check availability.
Recommended: DGX Spark, 128 GB unified, $0.79 / hour, billed per minute. Check availability.
Recommended: DGX Spark, 128 GB unified, $0.79 / hour, billed per minute. Check availability.
Configure
Every build below is published on Hugging Face and fits at least one GPU we rent. Pick the quantization and engine: the memory it needs, the GPUs it fits and the recommended GPU follow, and Deploy takes that exact build, revision and GPU to the console. Only the one-click preset is tested by GPUwerk.
| Quantization | AWQ 4-bit |
|---|---|
| Engine | vLLM |
| Status | One-click preset |
| Weights | 16.8 GB |
| GPU memory it needs | About 22 GB as the preset serves it |
| Recommended GPU | RTX 5090, $0.69 / hour |
| Weights source | QuantTrio/Qwen3-Coder-30B-A3B-Instruct-AWQ, revision c58857a7f41c |
| Licence on the card | apache-2.0 |
| GPU | Memory | Price | Availability | Deploy |
|---|---|---|---|---|
| RTX 5090x86 · Blackwell · recommended | 32 GB GDDR7 | $0.69 / hour | Check availability | Deploy now Qwen3 Coder 30B AWQ 4-bit on RTX 5090 |
| DGX SparkArm · GB10 | 128 GB unified | $0.79 / hour | Check availability | Deploy now Qwen3 Coder 30B AWQ 4-bit on DGX Spark |
| RTX PRO 6000x86 · Blackwell | 96 GB GDDR7 | $1.89 / hour | Check availability | Deploy now Qwen3 Coder 30B AWQ 4-bit on RTX PRO 6000 |
| H100 SXMx86 · Hopper | 80 GB HBM3 | $2.69 / hour | Check availability | Deploy now Qwen3 Coder 30B AWQ 4-bit on H100 SXM |
| H200x86 · Hopper | 141 GB HBM3e | Does not fit: the one-click preset does not run on this GPU. | ||
| B200x86 · Blackwell | 180 GB HBM3e | Does not fit: the one-click preset does not run on this GPU. | ||
| B300x86 · Blackwell | 288 GB HBM3e | Does not fit: the one-click preset does not run on this GPU. | ||
| Quantization | FP8 |
|---|---|
| Engine | vLLM |
| Status | Untested by GPUwerk |
| Weights | 31.2 GB |
| GPU memory it needs | About 38.9 GB with a 32k-token cache, an estimate |
| Recommended GPU | DGX Spark, $0.79 / hour |
| Weights source | Qwen/Qwen3-Coder-30B-A3B-Instruct-FP8, revision dcaee4d4dfc5 |
| Licence on the card | apache-2.0 |
| GPU | Memory | Price | Availability | Deploy |
|---|---|---|---|---|
| RTX 5090x86 · Blackwell | 32 GB GDDR7 | Does not fit: needs about 39 GB of GPU memory (weights 31.2 GB plus cache, an estimate); RTX 5090 offers 29 GB to it. | ||
| DGX SparkArm · GB10 · recommended | 128 GB unified | $0.79 / hour | Check availability | Run it yourself Qwen3 Coder 30B FP8 on DGX Spark |
| RTX PRO 6000x86 · Blackwell | 96 GB GDDR7 | $1.89 / hour | Check availability | Run it yourself Qwen3 Coder 30B FP8 on RTX PRO 6000 |
| H100 SXMx86 · Hopper | 80 GB HBM3 | $2.69 / hour | Check availability | Run it yourself Qwen3 Coder 30B FP8 on H100 SXM |
| H200x86 · Hopper | 141 GB HBM3e | $3.49 / hour | Check availability | Run it yourself Qwen3 Coder 30B FP8 on H200 |
| B200x86 · Blackwell | 180 GB HBM3e | $5.49 / hour | Check availability | Run it yourself Qwen3 Coder 30B FP8 on B200 |
| B300x86 · Blackwell | 288 GB HBM3e | $6.49 / hour | Check availability | Run it yourself Qwen3 Coder 30B FP8 on B300 |
| Quantization | BF16 |
|---|---|
| Engine | vLLM |
| Status | Untested by GPUwerk |
| Weights | 61.1 GB |
| GPU memory it needs | About 73.3 GB with a 32k-token cache, an estimate |
| Recommended GPU | DGX Spark, $0.79 / hour |
| Weights source | Qwen/Qwen3-Coder-30B-A3B-Instruct, revision b2cff646eb4b |
| Licence on the card | apache-2.0 |
| GPU | Memory | Price | Availability | Deploy |
|---|---|---|---|---|
| RTX 5090x86 · Blackwell | 32 GB GDDR7 | Does not fit: needs about 73 GB of GPU memory (weights 61.1 GB plus cache, an estimate); RTX 5090 offers 29 GB to it. | ||
| DGX SparkArm · GB10 · recommended | 128 GB unified | $0.79 / hour | Check availability | Run it yourself Qwen3 Coder 30B BF16 on DGX Spark |
| RTX PRO 6000x86 · Blackwell | 96 GB GDDR7 | $1.89 / hour | Check availability | Run it yourself Qwen3 Coder 30B BF16 on RTX PRO 6000 |
| H100 SXMx86 · Hopper | 80 GB HBM3 | Does not fit: needs about 73 GB of GPU memory (weights 61.1 GB plus cache, an estimate); H100 offers 72 GB to it. | ||
| H200x86 · Hopper | 141 GB HBM3e | $3.49 / hour | Check availability | Run it yourself Qwen3 Coder 30B BF16 on H200 |
| B200x86 · Blackwell | 180 GB HBM3e | $5.49 / hour | Check availability | Run it yourself Qwen3 Coder 30B BF16 on B200 |
| B300x86 · Blackwell | 288 GB HBM3e | $6.49 / hour | Check availability | Run it yourself Qwen3 Coder 30B BF16 on B300 |
| Quantization | Q4_K_M GGUF |
|---|---|
| Engine | llama.cpp |
| Status | Untested by GPUwerk |
| Weights | 18.6 GB |
| GPU memory it needs | About 24.4 GB with a 32k-token cache, an estimate |
| Recommended GPU | DGX Spark, $0.79 / hour |
| Weights source | unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF, revision b17cb02dd882, file Qwen3-Coder-30B-A3B-Instruct-Q4_K_M.gguf |
| Licence on the card | apache-2.0 |
| GPU | Memory | Price | Availability | Deploy |
|---|---|---|---|---|
| RTX 5090x86 · Blackwell | 32 GB GDDR7 | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| DGX SparkArm · GB10 · recommended | 128 GB unified | $0.79 / hour | Check availability | Run it yourself Qwen3 Coder 30B Q4_K_M GGUF on DGX Spark |
| RTX PRO 6000x86 · Blackwell | 96 GB GDDR7 | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| H100 SXMx86 · Hopper | 80 GB HBM3 | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| H200x86 · Hopper | 141 GB HBM3e | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| B200x86 · Blackwell | 180 GB HBM3e | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| B300x86 · Blackwell | 288 GB HBM3e | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| Quantization | Q8_0 GGUF |
|---|---|
| Engine | llama.cpp |
| Status | Untested by GPUwerk |
| Weights | 32.5 GB |
| GPU memory it needs | About 40.4 GB with a 32k-token cache, an estimate |
| Recommended GPU | DGX Spark, $0.79 / hour |
| Weights source | unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF, revision b17cb02dd882, file Qwen3-Coder-30B-A3B-Instruct-Q8_0.gguf |
| Licence on the card | apache-2.0 |
| GPU | Memory | Price | Availability | Deploy |
|---|---|---|---|---|
| RTX 5090x86 · Blackwell | 32 GB GDDR7 | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| DGX SparkArm · GB10 · recommended | 128 GB unified | $0.79 / hour | Check availability | Run it yourself Qwen3 Coder 30B Q8_0 GGUF on DGX Spark |
| RTX PRO 6000x86 · Blackwell | 96 GB GDDR7 | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| H100 SXMx86 · Hopper | 80 GB HBM3 | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| H200x86 · Hopper | 141 GB HBM3e | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| B200x86 · Blackwell | 180 GB HBM3e | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
| B300x86 · Blackwell | 288 GB HBM3e | Does not fit: GGUF models run on llama.cpp, which has no x86_64 build yet; pick a DGX Spark, or a safetensors build for vLLM. | ||
Builds by the model's maker, NVIDIA and Unsloth, read from Hugging Face on 21 November 2026. Memory is an estimate (weights x 1.15 plus 3 GB); GGUF runs on llama.cpp, which GPUwerk builds for the DGX Spark only, and NVFP4 needs a Blackwell GPU. SGLang is not offered: there is no SGLang runtime yet.
Details
About the model.
| Maker | Qwen |
|---|---|
| Task | Coding · Text |
| Parameters | 30B total, 3B active (mixture of experts) |
| Context | 32,768 tokens as served (one-click preset) |
| Licence | Apache 2.0 |
80.9 tok/s for one request on one DGX Spark, measured by GPUwerk with vLLM on 2 September 2026. Every other speed figure is in the model guide, with its source.
Quick start
- Deploy Qwen3 Coder 30B on an RTX 5090. The console opens with the GPU and model picked.
- Wait for the instance to show Running: the model server answers once the weights are loaded.
- Call the HTTPS endpoint from the instance page with an API key, like any OpenAI-compatible API.
curl https://<your-endpoint>/v1/chat/completions \
-H "Authorization: Bearer $GPUWERK_API_KEY" -H "Content-Type: application/json" \
-d '{"model": "qwen3-coder-30b", "messages": [{"role": "user", "content": "Hello"}]}'
SSH, the browser terminal, VNC and your workspace work the same on every instance. Terminate when you are done and billing ends.
Related models
Qwen3.6 35B-A3B
A fast mixture-of-experts model for chat and agents, also on the pay-per-token API.
GPT-OSS 120B
OpenAI's open-weight 120B model, quantized to fit one DGX Spark.
Qwen3.8 Flash Next
A large Qwen3.8 mixture-of-experts model, quantized to fit one DGX Spark.