Mistral
Codestral 2 25.08
前沿
Aug 2025发布日期256K tokens上下文MNPL许可证80 优秀质量
Codestral 2 25.08 (22B parameters) requires approximately 17.4 GB of VRAM with Q4_K_M quantization. For the best balance of quality and speed, we recommend hardware with at least 21 GB of VRAM.
快速开始
— 复制粘贴即可本地运行Copy-paste commands to run Codestral 2 25.08 on your machine.
Run
lms load codestral-2508 && lms server startQuick specs
Parameters22B
Architecturedense
Context256K tokens
Modalitytext
Min RAM8.6 GB
Rec. RAM13.4 GB (Q4_K_M)
LicenseMNPL
FamilyCodestral
✓ Code
About this model
相关模型
快速推荐
最佳硬件
Codestral 2 25.08 的最佳选择
运行此模型
量化选项
各量化级别的 VRAM 估算
No hardware detected — fit column shows raw VRAM estimates
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 8.6 GB | Low | — |
Q3_K_S | 3 | 10.8 GB | Low | — |
NVFP4 | 4 | 12.3 GB | Medium | — |
Q4_K_M | 4 | 13.4 GB | Medium | — |
Q5_K_M | 5 | 15.8 GB | High | — |
Q6_K | 6 | 18.0 GB | High | — |
Q8_0 | 8 | 23.5 GB | Very High | — |
F16 | 16 | 45.1 GB | Maximum | — |
Quality benchmarks
Codestral 2 25.08 benchmark scores
Coding
SWE-bench Verified—
HumanEval+86.6%
Aider Polyglot—
LiveCodeBench37.9%
Source: official · 2025-08-14
硬件兼容性
全部硬件的适配估算
Computing compatibility...
内存详细分析
Reference: RTX 2060 6GB
Weights13.4 GB
KV Cache2.4 GB
Runtime0.9 GB
Headroom0.6 GB
常见问题
FAQ — Codestral 2 25.08
另请参阅