Grok 4.7 · Writing

Grok 4.7 for writing: our test runs

Grok 4.7 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 free See pricing

Based on 5 of our test runs on October 2, 2026, through OpenRouter with the app's own prompt and settings. Updated October 15, 2026.

Short answer

In our test runs on October 2, 2026, Grok 4.7 passed 3 of 5 of our writing prompts (1 of the 2 hard ones): tied 16th of 20 models. Each reply counted as 1 message on Pro, where it has 250 a month, and cost $0.0058 to run on average.

Prompt by prompt

Each of the 5 writing prompts, and how Grok 4.7's reply did.

Grok 4.7 on each writing prompt
Prompt	Result	Cost	Time
Announce a second bakery shop on LinkedInEveryday	Passed: Graded 4.8 of 5 on average (lowest 4).	$0.0093	16.1 s
Rewrite corporate jargon in plain wordsEveryday	Failed: Graded 3.7 of 5 on average (lowest 3).	$0.0112	20.0 s
Decline a meeting and offer two timesEveryday	Passed: Graded 4.7 of 5 on average (lowest 4).	$0.0024	2.5 s
A product announcement with five rulesHard	Passed: Graded 4.0 of 5 on average (lowest 3).	$0.0025	2.5 s
Argue both sides of free busesHard	Failed: Graded 4.7 of 5 on average (lowest 4); but a paragraph of 100 words, over the 90 allowed.	$0.0034	6.0 s

Every writing reply from Grok 4.7

Announce a second bakery shop on LinkedIn

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

Passed: Graded 4.8 of 5 on average (lowest 4).

We're happy to tell you that Crumb & Co. is opening a second shop in Leeds on 14 November. Our small bakery has been shaped by the customers who drop in for fresh bread, a slice of cake, and a friendly word at the counter. That support means a great deal to us, and we're glad to bring the same simple welcome to a new street. Whether you already know us or this will be your first visit, you'll find the same ovens, the same recipes, and the same quiet care.

Opening day is 14 November, and coffee will be free for everyone who comes through the door. Please stop by, have a look around, and taste something straight from the oven. We would love to meet you there.

1,625 tokens in, 1,299 out (1,149 of them reasoning) · 16.1 s · $0.0093 · 1 message on Pro · answered by x-ai/grok-4.7 via xAI · October 2, 2026

Rewrite corporate jargon in plain words

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

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

In the third quarter, teams from different parts of the company worked together to put a customer-focused approach into practice. Retention measures rose 12%, and the rate of customer loss fell by 30 basis points among enterprise customers. Contact points across all channels were also resized to use capacity more efficiently.

1,658 tokens in, 1,603 out (1,543 of them reasoning) · 20.0 s · $0.0112 · 1 message on Pro · answered by x-ai/grok-4.7 via xAI · October 2, 2026

Decline a meeting and offer two times

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

Passed: Graded 4.7 of 5 on average (lowest 4).

Subject: Re: Vendor demo — Wednesday

Hi Tom,

Thanks for the invitation. I’m presenting to the board on Wednesday at 11:00, so I can’t make the vendor demo then.

Would Thursday at 10:00 or Friday at 14:00 work instead? Happy to join whichever is easier for you and the vendor.

Best regards

1,630 tokens in, 144 out (74 of them reasoning) · 2.5 s · $0.0024 · 1 message on Pro · answered by x-ai/grok-4.7 via xAI · October 2, 2026

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).
```
The Nordvik 2 desk lamp is ready for your workspace. It dims in five steps, charges from USB-C, and its arm turns 270 degrees. It costs €79 and ships on 1 December.

Will it earn a place on your desk?
```
1,642 tokens in, 155 out (102 of them reasoning) · 2.5 s · $0.0025 · 1 message on Pro · answered by x-ai/grok-4.7 via xAI · October 2, 2026

Argue both sides of free buses

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

Failed: Graded 4.7 of 5 on average (lowest 4); but a paragraph of 100 words, over the 90 allowed.

A city should make its buses free because fare collection costs money, slows boarding, and keeps poorer riders off the system. Zero fares would cut those overheads, fill empty seats, and turn transit into a genuine public service rather than a paywall. Riders who switch from cars would ease congestion and pollution, while workers, students, and elderly residents would gain reliable mobility without budgeting for every trip. The social return—fewer missed shifts, wider access to jobs and clinics, a more livable street—outweighs the lost ticket revenue if the city funds the gap from parking fees or a modest local tax.

Free buses fail when they are treated as a gift rather than a service that must still run on time. Without fares, demand surges while revenue collapses, so agencies cut frequency, defer maintenance, and pack vehicles until riding becomes slower and less safe than driving. Crowding and disorder drive away the very commuters whose fares once subsidized off-peak routes, and taxpayers who never ride absorb a growing bill with no accountability for performance. A cheap fare, paired with faster boarding and more buses, keeps the system solvent and trustworthy; abolishing the price does not create capacity, and a free bus that does not come is worthless.

1,619 tokens in, 307 out (54 of them reasoning) · 6.0 s · $0.0034 · 1 message on Pro · answered by x-ai/grok-4.7 via xAI · October 2, 2026

How it compares on writing

Grok 4.7 beside the job's picks among every model. 3 models share the hard-problems pick, and Grok 4.7 isn't one of them.

Grok 4.7 beside other models on writing
Model	Passed	On Pro	Cost per reply
DeepSeek V4 ProOur pick: hard problems (shared), best value	5 of 5	250 a month on Pro	$0.0011
GPT-6 LunaOur pick: hard problems (shared), everyday	5 of 5	60 a day on Pro	$0.0001
Grok 4.7	3 of 5	250 a month on Pro	$0.0058

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.

Grok 4.7 next to each model on writing

Grok 4.7 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: Grok 4.7 answered within 10% of its time (6.0 s to 6.0 s), cost 3.0× less, and passed 3 of 5 to its 3. Only Grok 4.7 passed A product announcement with five rules. Only Claude Fable 5.1 passed Rewrite corporate jargon in plain words.
Against Claude Opus 5.5: Grok 4.7 answered 1.4× sooner (6.0 s to 8.2 s), cost 3.2× less, and passed 3 of 5 to its 4. Only Claude Opus 5.5 passed Rewrite corporate jargon in plain words.
Against Claude Sonnet 5.5: Grok 4.7 answered 2.6× later (6.0 s to 2.3 s), cost 1.7× more, and passed 3 of 5 to its 4. Only Claude Sonnet 5.5 passed Rewrite corporate jargon in plain words.
Against Claude Sonnet 5: Grok 4.7 answered 1.3× later (6.0 s to 4.5 s), cost 1.7× more, and passed 3 of 5 to its 4. Only Grok 4.7 passed Announce a second bakery shop on LinkedIn. Only Claude Sonnet 5 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Claude Haiku 5.5: Grok 4.7 answered 3.9× later (6.0 s to 1.5 s), cost 35.7× more, and passed 3 of 5 to its 3.
Against Claude Haiku 4.5: Grok 4.7 answered 2.7× later (6.0 s to 2.2 s), cost 5.1× more, and passed 3 of 5 to its 4. Only Grok 4.7 passed A product announcement with five rules. Only Claude Haiku 4.5 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6 Astra: Grok 4.7 answered 1.2× later (6.0 s to 4.8 s), cost 1.9× less, and passed 3 of 5 to its 5. Only GPT-6 Astra passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6.1 Sol: Grok 4.7 answered 1.8× later (6.0 s to 3.3 s), cost 5.1× more, and passed 3 of 5 to its 4. Only GPT-6.1 Sol passed Argue both sides of free buses.
Against GPT-6 Sol: Grok 4.7 answered 1.7× later (6.0 s to 3.5 s), cost 2.5× more, and passed 3 of 5 to its 4. Only GPT-6 Sol passed Argue both sides of free buses.
Against GPT-6 Luna: Grok 4.7 answered 2.5× later (6.0 s to 2.4 s), cost 57.8× more, and passed 3 of 5 to its 5. Only GPT-6 Luna passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Gemini 4 Argon: Grok 4.7 answered within 10% of its time (6.0 s to 5.7 s), cost 4.1× more, and passed 3 of 5 to its 4. Only Gemini 4 Argon passed Rewrite corporate jargon in plain words.
Against Gemini 3.1 Pro (preview): Grok 4.7 answered 1.5× sooner (6.0 s to 9.1 s), cost 1.6× less, and passed 3 of 5 to its 4. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words.
Against Gemini 3.8 Flash: Grok 4.7 answered 1.5× later (6.0 s to 3.9 s), cost 4.4× more, and passed 3 of 5 to its 3. Only Grok 4.7 passed A product announcement with five rules. Only Gemini 3.8 Flash passed Argue both sides of free buses.
Against DeepSeek V4.1 Flash: Grok 4.7 answered 4.5× later (6.0 s to 1.3 s), cost 12.8× more, and passed 3 of 5 to its 4. Only DeepSeek V4.1 Flash passed Argue both sides of free buses.
Against DeepSeek V4 Pro: Grok 4.7 answered 1.5× later (6.0 s to 3.9 s), cost 5.4× more, and passed 3 of 5 to its 5. Only DeepSeek V4 Pro passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Kimi K3: Grok 4.7 answered 2.0× later (6.0 s to 3.0 s), cost 1.5× more, and passed 3 of 5 to its 4. Only Grok 4.7 passed A product announcement with five rules. Only Kimi K3 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GLM 5.3: Grok 4.7 answered 3.4× later (6.0 s to 1.8 s), cost 10.0× more, and passed 3 of 5 to its 4. Only GLM 5.3 passed Rewrite corporate jargon in plain words.
Against GLM 5.3 Flash: Grok 4.7 answered 1.4× sooner (6.0 s to 8.5 s), cost 18.0× more, and passed 3 of 5 to its 2. Only Grok 4.7 passed A product announcement with five rules.
Against Mistral Large 4: Grok 4.7 answered 1.6× sooner (6.0 s to 9.8 s), cost 2.4× more, and passed 3 of 5 to its 4. Only Grok 4.7 passed A product announcement with five rules. Only Mistral Large 4 passed Rewrite corporate jargon in plain words; 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.

Questions

Is Grok 4.7 good for writing?

In our test runs it passed 3 of 5 writing prompts, tied 16th 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 Grok 4.7 use?

1 message each on Pro, where it has 250 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.