Comprehensive side-by-side LLM comparison
DeepSeek-V2.5 leads with 14.8% higher average benchmark score. Gemma 3n E4B Instructed LiteRT Preview supports multimodal inputs. DeepSeek-V2.5 is available on 3 providers. Overall, DeepSeek-V2.5 is the stronger choice for coding tasks.
DeepSeek
DeepSeek-V2.5 was developed as an enhanced iteration of the DeepSeek-V2 architecture, designed to incorporate improvements in model quality and efficiency. Built to advance the DeepSeek foundation model series, it provides refined capabilities for general-purpose language understanding and generation tasks.
Gemma 3N E4B IT LiteRT Preview was introduced as an experimental LiteRT-optimized version, designed to showcase advanced on-device AI capabilities. Built to demonstrate the feasibility of running capable instruction-tuned models on mobile platforms, it represents the cutting edge of edge AI development.
1 year newer

DeepSeek-V2.5
DeepSeek
2024-05-08

Gemma 3n E4B Instructed LiteRT Preview
2025-05-20
Context window and performance specifications
Average performance across 2 common benchmarks

DeepSeek-V2.5

Gemma 3n E4B Instructed LiteRT Preview
Gemma 3n E4B Instructed LiteRT Preview
2024-06-01
Available providers and their performance metrics

DeepSeek-V2.5
DeepInfra
DeepSeek
Hyperbolic

DeepSeek-V2.5

Gemma 3n E4B Instructed LiteRT Preview

DeepSeek-V2.5

Gemma 3n E4B Instructed LiteRT Preview

Gemma 3n E4B Instructed LiteRT Preview