Qwen2.5-3B features 3 billion parameters, 36 layers, and a 16/2 attention head architecture, delivering a significant performance leap while maintaining reasonable resource requirements. The model supports a 32K-token context window and up to 8K-token generation, enabling it to handle moderately complex tasks with extended contexts.
The uniqueness of Qwen2.5-3B lies in its return to the product line after being absent from the Qwen2 series, effectively filling the crucial gap between 1.5B and 7B models. This size proves particularly valuable for resource-constrained scenarios where the 7B version might be excessive, yet higher performance than the 1.5B variant is required. The model demonstrates substantially improved capabilities in understanding complex instructions, multi-step reasoning, and working with structured data.
Notably, this model is distributed under the Qwen Research License, which may impose certain restrictions on commercial use. However, Qwen2.5-3B is ideally suited for research projects, prototyping, and developing specialized solutions where licensing flexibility for research purposes is essential. The model performs exceptionally well in data analysis tasks, technical documentation processing, educational applications, and serves as an excellent base for creating domain-specific models through fine-tuning.
| Model Name | Context | Type | GPU | Status | Link |
|---|
There are no public endpoints for this model yet.
Rent your own physically dedicated instance with hourly or long-term monthly billing.
We recommend deploying private instances in the following scenarios:
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
32,768.0 |
1 | $0.33 | 7.262 | Launch | ||
32,768.0 |
1 | $0.38 | 3.991 | Launch | ||
32,768.0 |
1 | $0.38 | 7.294 | Launch | ||
32,768.0 |
1 | $0.53 | 13.755 | Launch | ||
32,768.0 |
1 | $0.57 | 3.214 | Launch | ||
32,768.0 |
1 | $0.83 | 14.622 | Launch | ||
32,768.0 |
1 | $1.02 | 14.589 | Launch | ||
32,768.0 tensor |
2 | $1.23 | 29.222 | Launch | ||
32,768.0 |
1 | $1.59 | 20.992 | Launch | ||
32,768.0 |
1 | $2.37 | 60.164 | Launch | ||
32,768.0 |
1 | $3.83 | 60.107 | Launch | ||
32,768.0 |
1 | $4.11 | 71.482 | Launch | ||
32,768.0 tensor |
2 | $4.61 | 122.039 | Launch | ||
32,768.0 |
1 | $4.74 | 109.680 | Launch | ||
32,768.0 tensor |
2 | $9.40 | 221.072 | Launch | ||
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
32,768.0 |
1 | $0.33 | 6.104 | Launch | ||
32,768.0 |
1 | $0.38 | 2.833 | Launch | ||
32,768.0 |
1 | $0.38 | 42.540 | 6.137 | Launch | |
32,768.0 |
1 | $0.53 | 93.330 | 12.597 | Launch | |
32,768.0 |
1 | $0.57 | 2.056 | Launch | ||
32,768.0 |
1 | $0.83 | 13.464 | Launch | ||
32,768.0 |
1 | $1.02 | 13.432 | Launch | ||
32,768.0 tensor |
2 | $1.23 | 28.064 | Launch | ||
32,768.0 |
1 | $1.59 | 19.835 | Launch | ||
32,768.0 |
1 | $2.37 | 156.450 | 59.006 | Launch | |
32,768.0 |
1 | $3.83 | 137.720 | 58.949 | Launch | |
32,768.0 |
1 | $4.11 | 187.370 | 70.324 | Launch | |
32,768.0 tensor |
2 | $4.61 | 120.882 | Launch | ||
32,768.0 |
1 | $4.74 | 108.523 | Launch | ||
32,768.0 tensor |
2 | $9.40 | 219.915 | Launch | ||
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
32,768.0 |
1 | $0.33 | 3.864 | Launch | ||
32,768.0 |
1 | $0.38 | 3.897 | Launch | ||
32,768.0 |
1 | $0.53 | 10.357 | Launch | ||
32,768.0 tensor |
2 | $0.69 | 6.295 | Launch | ||
32,768.0 |
1 | $0.83 | 11.224 | Launch | ||
32,768.0 tensor |
2 | $0.97 | 4.742 | Launch | ||
32,768.0 |
1 | $1.02 | 11.191 | Launch | ||
32,768.0 tensor |
2 | $1.23 | 25.824 | Launch | ||
32,768.0 |
1 | $1.59 | 17.595 | Launch | ||
32,768.0 |
1 | $2.37 | 56.766 | Launch | ||
32,768.0 |
1 | $3.83 | 56.709 | Launch | ||
32,768.0 |
1 | $4.11 | 68.084 | Launch | ||
32,768.0 tensor |
2 | $4.61 | 118.642 | Launch | ||
32,768.0 |
1 | $4.74 | 106.283 | Launch | ||
32,768.0 tensor |
2 | $9.40 | 217.675 | Launch | ||
Contact our dedicated neural networks support team at nn@immers.cloud or send your request to the sales department at sale@immers.cloud.