Skip to content
llmwise
Sign inStart free

Gemini 3.1 Pro · Writing

Gemini 3.1 Pro for writing: our test runs

Gemini 3.1 Pro ran our 5 writing prompts in llmwise, through the same pipeline your messages take. Here's every reply as it came, whether it passed, what it cost and what it counts as on Pro, next to the other models.

Start freeSee pricing

Based on 5 of our test runs on , through OpenRouter with the app's own prompt and settings. Updated .

Short answer

In our test runs on September 27, 2026, Gemini 3.1 Pro passed 4 of 5 of our writing prompts (1 of the 2 hard ones): tied 4th of 20 models. Each reply counted as 1 message on Pro, where it has 125 a month, and cost $0.0092 to run on average.

Prompt by prompt

Each of the 5 writing prompts, and how Gemini 3.1 Pro's reply did.

Gemini 3.1 Pro on each writing prompt
PromptResultCostTime
Announce a second bakery shop on LinkedInEverydayPassed: Graded 4.0 of 5 on average (lowest 3).$0.01139.5 s
Rewrite corporate jargon in plain wordsEverydayPassed: Graded 4.0 of 5 on average (lowest 3).$0.01018.8 s
Decline a meeting and offer two timesEverydayPassed: Graded 5.0 of 5 on average (lowest 5).$0.00586.2 s
A product announcement with five rulesHardPassed: Graded 4.0 of 5 on average (lowest 3).$0.00959.2 s
Argue both sides of free busesHardFailed: Graded 3.7 of 5 on average (lowest 3).$0.00929.1 s

Every writing reply from Gemini 3.1 Pro

  1. Announce a second bakery shop on LinkedIn

    An everyday prompt: “Announce a second bakery shop on LinkedIn”, in full.

    Passed: Graded 4.0 of 5 on average (lowest 3).

    We are absolutely thrilled to share some wonderful news with you all. Our little bakery, Crumb & Co., is growing! Thanks to your incredible support over the years, we are opening the doors to our second shop in Leeds on the 14th of November. 

When we first started baking, we never imagined we would be able to share our pastries and breads with so many lovely people. This new chapter in Leeds means the world to us, and we cannot wait to become a part of the local community. 

To celebrate this exciting milestone, we would love for you to join us on our opening day. Pop in to say hello, check out the new space, and enjoy a free cup of coffee on us. We are so grateful for your continued support and look forward to welcoming you to our new home very soon.

    417 tokens in, 876 out (707 of them reasoning) · 9.5 s · $0.0113 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

  2. Rewrite corporate jargon in plain words

    An everyday prompt: “Rewrite corporate jargon in plain words”, in full.

    Passed: Graded 4.0 of 5 on average (lowest 3).

    In Q3, our teams worked together to adopt a more customer-focused approach. For our business clients, this resulted in a 12% increase in customer retention and a 0.30% (30 basis points) decrease in the rate of customers leaving. Additionally, we adjusted our various customer communication channels to make more efficient use of our time and resources.

    447 tokens in, 763 out (687 of them reasoning) · 8.8 s · $0.0101 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

  3. Decline a meeting and offer two times

    An everyday prompt: “Decline a meeting and offer two times”, in full.

    Passed: Graded 5.0 of 5 on average (lowest 5).

    Subject: Re: Vendor Demo

Hi Tom,

Thank you for the invitation to the vendor demo on Wednesday at 11:00. Unfortunately, I won't be able to make it because I am presenting to the board at that time. 

Could we reschedule for later in the week? I am available on Thursday at 10:00 or Friday at 14:00. Let me know if either of those times works for you.

Best,

[Your Name]

    428 tokens in, 415 out (310 of them reasoning) · 6.2 s · $0.0058 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

  4. A product announcement with five rules

    A hard prompt: “A product announcement with five rules”, in full.

    Passed: Graded 4.0 of 5 on average (lowest 3).

    Introducing the Nordvik 2 desk lamp, designed to illuminate your workspace with precision. This new model features a highly flexible arm that turns 270 degrees, putting light exactly where you need it. You can easily tailor the brightness to suit any task, as it dims in five steps. Built for modern convenience, it also charges from USB-C to keep your setup clean and wire-free. The Nordvik 2 costs €79 and ships on 1 December. Are you ready to upgrade your daily lighting?

    435 tokens in, 715 out (609 of them reasoning) · 9.2 s · $0.0095 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

  5. Argue both sides of free buses

    A hard prompt: “Argue both sides of free buses”, in full.

    Failed: Graded 3.7 of 5 on average (lowest 3).

    Making city buses free is a powerful catalyst for economic equality and environmental sustainability. By eliminating fares, cities immediately remove a significant financial burden from low-income residents, improving their access to jobs, education, and healthcare. Furthermore, free transit incentivizes people to leave their cars at home, dramatically reducing traffic congestion and cutting harmful greenhouse gas emissions. This shift not only improves urban air quality but also reduces the long-term wear and tear on city infrastructure, ultimately creating a cleaner, fairer, and more accessible environment for all residents.

However, eliminating fares fundamentally undermines the financial stability and quality of the transit system itself. Fares provide crucial revenue required for essential maintenance, route expansion, and driver salaries. Without this income, cities are forced to either slash service frequencies and coverage or drastically raise local taxes to cover the deficit. Consequently, free transit often results in overcrowded, unreliable, and dilapidated buses. Passengers ultimately prioritize speed, safety, and reliability over zero cost; a free bus is useless to a commuter if it never arrives on time.

    409 tokens in, 699 out (492 of them reasoning) · 9.1 s · $0.0092 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

How it compares on writing

Gemini 3.1 Pro beside the job's picks among every model. 3 models share the hard-problems pick, and Gemini 3.1 Pro isn't one of them.

Gemini 3.1 Pro beside other models on writing
ModelPassedOn ProCost per reply
DeepSeek V4 ProOur pick: hard problems (shared), best value5 of 5250 a month on Pro$0.0011
GPT-6 LunaOur pick: hard problems (shared), everyday5 of 560 a day on Pro$0.0001
Gemini 3.1 Pro4 of 5125 a month on Pro$0.0092
On Pro: Pro's count on each model. Cost: what OpenRouter charged us per reply, on average.

Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Plus $3.75, Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.

Gemini 3.1 Pro next to each model on writing

Gemini 3.1 Pro against each model on the same 5 writing prompts. One run each, through OpenRouter: a wait depends on the provider and the load that day, so a lead under 10% counts as close.

  • Against Claude Fable 5.1: Gemini 3.1 Pro (preview) answered 1.5× later (9.1 s to 6.0 s), cost 1.9× less, and passed 4 of 5 to its 3. Only Gemini 3.1 Pro (preview) passed A product announcement with five rules.

  • Against Claude Opus 5.5: Gemini 3.1 Pro (preview) answered 1.1× later (9.1 s to 8.2 s), cost 2.0× less, and passed 4 of 5 to its 4.

  • Against Claude Sonnet 5.5: Gemini 3.1 Pro (preview) answered 3.9× later (9.1 s to 2.3 s), cost 2.7× more, and passed 4 of 5 to its 4.

  • Against Claude Sonnet 5: Gemini 3.1 Pro (preview) answered 2.0× later (9.1 s to 4.5 s), cost 2.7× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed Announce a second bakery shop on LinkedIn. Only Claude Sonnet 5 passed Argue both sides of free buses.

  • Against Claude Haiku 5.5: Gemini 3.1 Pro (preview) answered 6.0× later (9.1 s to 1.5 s), cost 57.0× more, and passed 4 of 5 to its 3. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words.

  • Against Claude Haiku 4.5: Gemini 3.1 Pro (preview) answered 4.1× later (9.1 s to 2.2 s), cost 8.1× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed A product announcement with five rules. Only Claude Haiku 4.5 passed Argue both sides of free buses.

  • Against GPT-6 Astra: Gemini 3.1 Pro (preview) answered 1.9× later (9.1 s to 4.8 s), cost 1.2× less, and passed 4 of 5 to its 5. Only GPT-6 Astra passed Argue both sides of free buses.

  • Against GPT-6.1 Sol: Gemini 3.1 Pro (preview) answered 2.8× later (9.1 s to 3.3 s), cost 8.1× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words. Only GPT-6.1 Sol passed Argue both sides of free buses.

  • Against GPT-6 Sol: Gemini 3.1 Pro (preview) answered 2.6× later (9.1 s to 3.5 s), cost 4.1× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words. Only GPT-6 Sol passed Argue both sides of free buses.

  • Against GPT-6 Luna: Gemini 3.1 Pro (preview) answered 3.8× later (9.1 s to 2.4 s), cost 92.3× more, and passed 4 of 5 to its 5. Only GPT-6 Luna passed Argue both sides of free buses.

  • Against Gemini 4 Argon: Gemini 3.1 Pro (preview) answered 1.6× later (9.1 s to 5.7 s), cost 6.6× more, and passed 4 of 5 to its 4.

  • Against Gemini 3.8 Flash: Gemini 3.1 Pro (preview) answered 2.4× later (9.1 s to 3.9 s), cost 6.9× more, and passed 4 of 5 to its 3. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words; A product announcement with five rules. Only Gemini 3.8 Flash passed Argue both sides of free buses.

  • Against DeepSeek V4.1 Flash: Gemini 3.1 Pro (preview) answered 6.8× later (9.1 s to 1.3 s), cost 20.4× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words. Only DeepSeek V4.1 Flash passed Argue both sides of free buses.

  • Against DeepSeek V4 Pro: Gemini 3.1 Pro (preview) answered 2.3× later (9.1 s to 3.9 s), cost 8.6× more, and passed 4 of 5 to its 5. Only DeepSeek V4 Pro passed Argue both sides of free buses.

  • Against Grok 4.7: Gemini 3.1 Pro (preview) answered 1.5× later (9.1 s to 6.0 s), cost 1.6× more, and passed 4 of 5 to its 3. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words.

  • Against Kimi K3: Gemini 3.1 Pro (preview) answered 3.1× later (9.1 s to 3.0 s), cost 2.4× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed A product announcement with five rules. Only Kimi K3 passed Argue both sides of free buses.

  • Against GLM 5.3: Gemini 3.1 Pro (preview) answered 5.2× later (9.1 s to 1.8 s), cost 16.0× more, and passed 4 of 5 to its 4.

  • Against GLM 5.3 Flash: Gemini 3.1 Pro (preview) answered within 10% of its time (9.1 s to 8.5 s), cost 28.7× more, and passed 4 of 5 to its 2. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words; A product announcement with five rules.

  • Against Mistral Large 4: Gemini 3.1 Pro (preview) answered within 10% of its time (9.1 s to 9.8 s), cost 3.8× more, and passed 4 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed A product announcement with five rules. Only Mistral Large 4 passed Argue both sides of free buses.

How these runs were done

Rubric (graded). The grader model scores the reply from 1 to 5 on each published criterion. It passes with an average of 4 or more and no criterion under 3, and only if it also meets the prompt's automatic rules (length, words it must or mustn't use).

The grader is Claude Opus 5.5 at low reasoning effort; its own replies are graded by GPT-6 Astra, so no model grades itself. Its prompt and every rubric are on the methods page.

How the runs were done, and every writing prompt.

Gemini 3.1 Pro, and writing, elsewhere

Questions

Is Gemini 3.1 Pro good for writing?

In our test runs it passed 4 of 5 writing prompts, tied 4th of the 20 models in llmwise. Every reply is on this page, so you can judge them yourself.

How many of my messages does a writing reply from Gemini 3.1 Pro use?

1 message each on Pro, where it has 125 a month on Pro. The price of a message is fixed and shown before you send it, however long the reply.

How were these runs done?

The same way for every model: each prompt sent through llmwise's own pipeline, each reply checked the same way. The methods page has every prompt and how each is scored.

Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral, in one chat.

See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.

Start freeSee pricing