Qwen3.8 27B vs Gemma 4 31B.

A careful Qwen against Google's budget writer and image reader.

These are the mid-size options: small enough to answer quickly, large enough to think. Qwen3.8 27B is Alibaba's newest mid-size model, tuned for coding and careful answers; Gemma 4 31B is Google's open model, tuned for writing and reading images.

The price gap is the story. Qwen3.8 27B is listed at $3.20 per million tokens, while Gemma 4 31B is $0.40 — eight times cheaper. Both have roughly a quarter-million-token window, both accept images, use tools and offer zero data retention. Qwen adds a low default reasoning effort, so it tends to think before answering.

At a glance

Where Qwen3.8 27B pulls ahead

• Coding and careful answers: the model is explicitly tuned for code and for thinking through a problem before replying.
• Tasks where a wrong answer is expensive — migrations, tricky refactors, multi-step analysis — justify paying for the extra deliberation.
• A newer Qwen release with a reasoning-first default, which suits debugging and specification work.

Where Gemma 4 31B pulls ahead

• Price: at $0.40 per million tokens it is one eighth the upstream cost of Qwen3.8 27B.
• Writing and editing: Google tunes Gemma for prose, summaries and rewrites, and it reads images as well as text.
• High-volume work where eight times the price cannot be recovered by a small quality gain.

The price difference

Qwen3.8 27B costs $3.20 per million tokens against $0.40 for Gemma 4 31B — about 8.0x more. On UsingOpen you can chat with both for free inside the daily limit, so the rate is a fair proxy for how expensive each model is to run rather than a bill you pay here. Multiply it by your monthly token volume if you are comparing API access elsewhere.

Which should you choose?

Gemma 4 31B is the value pick and the better default for writing, summarising and image questions. It is hard to justify eight times the price for those jobs when the window and capabilities are so close.

Qwen3.8 27B earns its price only on code and reasoning where care matters: a difficult refactor, a migration, a long analysis. Test the exact task in Compare mode before committing; if the answers are close, keep the cheaper model.

Chat with Qwen3.8 27B free · Chat with Gemma 4 31B free

Frequently asked questions

Which has the larger context window, Qwen3.8 27B or Gemma 4 31B?

Qwen3.8 27B is listed at 262k (200,000 words); Gemma 4 31B is listed at 256k (190,000 words). Qwen3.8 27B can hold more in one conversation, which matters for long codebases, archives and multi-hour agent runs.

Can both Qwen3.8 27B and Gemma 4 31B read images?

Yes. Both models accept image input, so a screenshot or diagram is fair game in either. The decision should turn on context window, price, retention terms and how each model answers your own prompts.

Is Qwen3.8 27B or Gemma 4 31B better for coding?

Qwen3.8 27B is the closer fit for code here — the catalog tunes it for coding and careful answers — while Gemma 4 31B is aimed at writing, and reading images. For project-scale coding, start with Qwen3.8 27B; if the task leans toward the other model's strengths, it may still win.

Which is cheaper, Qwen3.8 27B or Gemma 4 31B?

Gemma 4 31B is listed at $0.40 per million tokens against $3.20 for Qwen3.8 27B — about 8.0x cheaper. Both are free to chat with on UsingOpen inside the daily limit when live; the rates matter if you are comparing what API access would cost.

Can I chat with Qwen3.8 27B and Gemma 4 31B for free?

Yes. Both are selectable on UsingOpen without an account when they are live, and each user gets 10 free prompts per day. A model paused for capacity is unavailable until it returns; the picker shows live status. History stays on your device, and Compare mode can send one prompt to both models side by side.

Why is Qwen3.8 27B eight times the price of Gemma 4 31B?

Size is not the reason — 27B versus 31B is nearly a wash. The makers price their models differently, and Qwen3.8 27B is positioned as a reasoning and coding model with a low default thinking effort. Whether that is worth 8x depends entirely on how much you value the answers on your own tasks.

Is Gemma 4 31B good enough for coding?

It can write and explain code, and it has tool support, but it is tuned primarily for writing and image reading. Qwen3.8 27B is the coding-focused option of the pair. For occasional snippets Gemma is fine; for project-scale work, compare them on your actual code first.

Related comparisons

• Qwen3.8 Flash vs Gemma 4 26B — Two fast multimodal models, one with a million-token window and one with zero-retention terms.
• GLM 5.3 Flash vs DeepSeek V4 Flash — The site's default all-rounder against the cheapest heavy hitter.
• Nemotron 3.5 Lightning vs Ling 3.0 Flash VL — NVIDIA's fast text specialist against inclusionAI's cheap multimodal tool user.

All models · Compare models · Free AI chat · FAQ