Layout1 gives you access to Qwen for your coding projects, through Claude Code and compatible tools.
Invite-only pilot. Joining the waitlist does not grant immediate access.
Your data is handled in accordance with our privacy policy.
The first answer is rarely the last. Give yourself room to compare approaches and refine your solution.
A stubborn bug, a refactor, a test suite. Work with Qwen on tasks that take more than one exchange.
Volume is only part of the picture. Explore Qwen’s evaluations to see what it can do.
Explore the evaluationsPut Qwen to the test on your own code. Access to the Layout1 pilot is by invitation from the waitlist.
Connect your coding tool to Layout1, then use Qwen without limits or counting tokens.
Open your workspace on the platform to access the setup for your tool.
Link your usual tool to Layout1 to use Qwen directly in your projects.
Open your project and use Qwen in your usual tool, with no token quota to keep track of.
Use Qwen through Layout1 in Claude Code or a compatible editor. Connection settings depend on the tool you choose.
Qwen3.8-Flash-Next compared with Claude Opus 4.8, GPT-5.5 and Gemini 3.1 Pro Preview.
Five tests selected to highlight its strengths against versions released from February to May 2026. This selection is not an overall ranking.
Sources checked on 2026-11-02
Produce analyses, presentations and documents — overall Elo score · Artificial Analysis · Opus / Qwen · GPT / Qwen · Gemini / Qwen
Elo scores · scale from 0 to 2,000
Qwen3.8-Flash-NextReasoning effort: Not reported
Claude Opus 4.8Reasoning effort: max
GPT-5.5Reasoning effort: xhigh
Gemini 3.1 Pro PreviewReasoning effort: Not reported
Complete tasks in a terminal · Artificial Analysis · Opus / Qwen · GPT / Qwen · Gemini / Qwen
Scores in % · scale from 0 to 100
Qwen3.8-Flash-NextReasoning effort: Not reported
Claude Opus 4.8Reasoning effort: max
GPT-5.5Reasoning effort: xhigh
Gemini 3.1 Pro PreviewReasoning effort: Not reported
Complete objectives across applications while following safeguards · Artificial Analysis · Opus / Qwen · GPT / Qwen · Gemini / Qwen
Scores in % · scale from 0 to 100
Qwen3.8-Flash-NextReasoning effort: Not reported
Claude Opus 4.8Reasoning effort: max
GPT-5.5Reasoning effort: xhigh
Gemini 3.1 Pro PreviewReasoning effort: Not reported
Resolve banking requests using tools — pass@1 · Artificial Analysis · Opus / Qwen · GPT / Qwen · Gemini / Qwen
Scores in % · scale from 0 to 100
Qwen3.8-Flash-NextReasoning effort: Not reported
Claude Opus 4.8Reasoning effort: max
GPT-5.5Reasoning effort: xhigh
Gemini 3.1 Pro PreviewReasoning effort: Not reported
Complete professional tasks — Elo score · Artificial Analysis · Opus / Qwen · GPT / Qwen · Gemini / Qwen
Elo scores · scale from 0 to 2,000
Qwen3.8-Flash-NextReasoning effort: Not reported
Claude Opus 4.8Reasoning effort: max
GPT-5.5Reasoning effort: xhigh
Gemini 3.1 Pro PreviewReasoning effort: Not reported
Results for practical tasks.
Coding, automation and document work: independent Artificial Analysis evaluations, with the versions and settings shown. These results do not measure our service’s performance.
What to know about access, compatible tools and the pilot.
Want to try Layout1? Join the waitlist.
