What is the best local model for coding on the RTX 3080 + RTX 5080?

On the RTX 3080 + RTX 5080, huihui-swift15-qwen3.8-27b ranks first for coding in our benchmarks (79.7/100 at ~120.7 tok/s).

What is the best local model for agentic workflows on the RTX 3080 + RTX 5080?

On the RTX 3080 + RTX 5080, huihui-swift15-qwen3.8-27b ranks first for agentic workflows in our benchmarks (83.0/100 at ~120.7 tok/s).

Hardware Benchmarks

Compare RTX 3080 + RTX 5080 benchmark results and see which models actually earn the best completed quality scores.

GPU-specific specifications for local LLM planning.

NVIDIA GeForce RTX 3080 + NVIDIA GeForce RTX 5080

Models we recommend for RTX 3080 + RTX 5080.

Fastest & Best quality & Largest context

27B

134.0 tok/sQuality 88.8Context 65,536 tokens

Highest token generation speed at 134.0 tok/s. Highest completed quality score at 88.8. Largest context that fits in 26GB VRAM, at 65,536 tokens.

Which models actually run best on the RTX 3080 + RTX 5080, by task, from community benchmark data.

What is the best local model for coding on the RTX 3080 + RTX 5080?: On the RTX 3080 + RTX 5080, huihui-swift15-qwen3.8-27b ranks first for coding in our benchmarks (79.7/100 at ~120.7 tok/s).
What is the best local model for agentic workflows on the RTX 3080 + RTX 5080?: On the RTX 3080 + RTX 5080, huihui-swift15-qwen3.8-27b ranks first for agentic workflows in our benchmarks (83.0/100 at ~120.7 tok/s).

Completed quality-ranked benchmark results for NVIDIA GeForce RTX 3080 + NVIDIA GeForce RTX 5080.