Model comparison
Poolside Laguna XS 2.1 (free) vs GPT-5
Capability radar, value bars, and side-by-side posture.
Poolside Laguna XS 2.1 (free)
Not scored
value score
GPT-5
94
value score
Poolside Laguna XS 2.1 (free) ctx
—
GPT-5 ctx
400K
Capability radar
Value · context · multimodal · openness · speed posture
Value head-to-head
- GPT-594
Attribute tape
| Field | Poolside Laguna XS 2.1 (free) | GPT-5 |
|---|---|---|
| Developer | poolside | OpenAI |
| Context | — | 400K |
| Modalities | Text, Code, Image, Audio | |
| Openness | — | Proprietary Model |
| Speed | — | Fast |
| Price | — | Mid-High |
| Value score | — | 94 |
| Summary | Laguna XS 2.1 is a 33B Mixture-of-Experts coding-agent model from Poolside, with 3B activated parameters per token and a 262K-token context window. It is optimized for agentic software engineering workflows including code generation, repository-scale reasoning, and terminal-style tool use, and is available as a free inference endpoint on OpenRouter. | OpenAI frontier multimodal model. |