Best model for your job
Best local LLM for coding
For local coding, Kimi K3 scores highest on the current data (66.0), weighting coding 70%, instruction following 30%. The cheapest model in the top 10 is MiMo-V2.6-Flash at $0.14 / $0.28 per million tokens.
Last verified
Weights: Coding 70%, Instruction Following 30%. Only models with downloadable weights qualify.
| # | Model | Provider | Fit | Coding | Instruction Following | Input $/M | Output $/M |
|---|---|---|---|---|---|---|---|
| 1 | Kimi K3 (open weights) | Moonshot AI | 66.0 | 61.0 | 77.7 | $3 | $15 |
| 2 | GLM-5.3 (open weights) | Z.ai (Zhipu) | 64.9 | 59.5 | 77.5 | $1.40 | $4.40 |
| 3 | MiMo-V2.6-Pro (open weights) | Xiaomi | 62.3 | 55.5 | 78.2 | $0.43 | $0.87 |
| 4 | MiMo-V2.6-Flash (open weights) | Xiaomi | 60.5 | 53.4 | 76.8 | $0.14 | $0.28 |
| 5 | GLM-5.3-Flash (open weights) | Z.ai (Zhipu) | 60.4 | 53.1 | 77.5 | $0.15 | $0.50 |
| 6 | DeepSeek V4.1 Flash (open weights) | DeepSeek | 60.2 | 52.9 | 77.3 | $0.15 | $0.60 |
| 7 | DeepSeek V4 Pro (open weights) | DeepSeek | 59.5 | 52.4 | 76.1 | $0.66 | $1.98 |
| 8 | GLM-5.2 (open weights) | Z.ai (Zhipu) | 59.0 | 51.3 | 76.9 | $1.40 | $4.40 |
| 9 | Kimi K2.6 (open weights) | Moonshot AI | 58.4 | 50.7 | 76.3 | $0.95 | $4 |
| 10 | Qwen3.8 27B (open weights) | Alibaba (Qwen) | 58.1 | 50.5 | 75.8 | $0.99 | $1.49 |
| 11 | GLM-5.1 (open weights) | Z.ai (Zhipu) | 57.0 | 48.7 | 76.3 | $1.40 | $4.40 |
| 12 | GLM-5 (open weights) | Z.ai (Zhipu) | 56.9 | 49.0 | 75.2 | $1 | $3.20 |
| 13 | Kimi K2.5 (open weights) | Moonshot AI | 56.8 | 48.8 | 75.3 | $0.45 | $2.25 |
| 14 | MiMo-V2.5-Pro (open weights) | Xiaomi | 56.4 | 47.4 | 77.5 | $0.43 | $0.87 |
| 15 | DeepSeek V4 Flash (open weights) | DeepSeek | 56.0 | 47.9 | 74.9 | $0.15 | $0.60 |
| 16 | Hy3 (open weights) | Tencent | 55.3 | 46.8 | 75.1 | $0.0825 | $0.33 |
| 17 | MiniMax-M2.5 (open weights) | MiniMax | 55.1 | 48.1 | 71.5 | $0.30 | $1.20 |
| 18 | DeepSeek-V3.2-Exp (open weights) | DeepSeek | 54.9 | 46.5 | 74.5 | $0.26 | $0.38 |
| 19 | MiMo-V2.5 (open weights) | Xiaomi | 53.4 | 43.9 | 75.5 | $0.14 | $0.28 |
| 20 | GLM-4.7 (open weights) | Z.ai (Zhipu) | 53.1 | 44.0 | 74.4 | $0.60 | $2.20 |
| 21 | Longcat Flash Chat (open weights) | Meituan | 52.8 | 43.5 | 74.4 | — | — |
| 22 | Qwen3 235B-A22B (open weights) | Alibaba (Qwen) | 52.8 | 44.3 | 72.6 | $0.70 | $2.80 |
| 23 | Inkling-Small (open weights) | Thinking Machines Lab | 52.7 | 43.6 | 73.8 | $0.45 | $1.20 |
| 24 | Kimi K2.5 Instant (open weights) | Moonshot AI | 52.4 | 42.6 | 75.3 | — | — |
| 25 | Gemma 4 31B IT (open weights) | 52.2 | 42.3 | 75.5 | $0.09 | $0.34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Top two head to head: Kimi K3 vs GLM-5.3
Frequently asked questions
What is the best local llm for coding?
For local coding, Kimi K3 scores highest on the current data (66.0), weighting coding 70%, instruction following 30%. The cheapest model in the top 10 is MiMo-V2.6-Flash at $0.14 / $0.28 per million tokens.
How is this shortlist built?
Only models with downloadable weights qualify. Each model's category scores are blended with those weights; only ranked models with results in every needed category are listed.
Other jobs
- Best AI model for coding
- Best AI model for debugging
- Best AI model for building websites
- Best AI model for agents
- Best AI model for RAG
- Best AI model for research
- Best AI model for studying
- Best AI model for math
- Best AI model for hard reasoning
- Best AI model for data analysis
- Best AI model for Excel
- Best AI model for data extraction
- Best AI model for writing
- Best AI model for emails
- Best AI model for copywriting
- Best AI model for creative writing
- Best AI model for translation
- Best AI model for language learning
- Best AI model for customer service
- Best AI model for presentations
- Best AI model for PDFs
- Best AI assistant for everyday use