The short answer: what a book translation costs
An 80,000-word novel translated from English to Spanish at standard commercial quality plans at 6,400 dollars to 9,600 dollars for the translation itself, plus about 15 percent for an independent proofreading pass, which lands near 7,360 dollars to 11,040 dollars all in. That single sentence answers the question most authors type into search, and everything below explains how to adjust it for your word count, your language direction, your quality tier, and your sales outlook. Harder language directions cost more per word, premium literary handling costs more than standard commercial handling, and a separate proofing pass adds a fixed percentage on top. The last step is division: take the planning total and divide it by the number of translated-edition copies you honestly expect to sell in the first two years. That per-sale figure tells you whether the project is comfortable, tight, or too heavy to approve on those sales alone.
Use the calculator at the top of this page to run your own numbers before you read further. Enter the final manuscript word count, pick the language direction, choose standard or premium quality, leave proofing switched on unless a vendor bundles it in writing, and enter realistic expected sales for the translated edition. The tool returns a low-to-high translation range, a low-to-high proofing range, a combined planning total, and a cost-per-expected-sale range. Treat every figure as a planning band for budgeting, not as a vendor quote and not as a promise of market performance. Real quotes move with translator experience, subject difficulty, schedule pressure, file quality, and the amount of editing a vendor includes, so your job is to budget with bands first and then confirm with two or three written quotes.
How the tool math works, step by step
The arithmetic has four stages and nothing hidden. First it looks up the per-word planning band for your language direction and quality tier. Second it multiplies that band by your billable word count to get the translation range. Third it applies the proofing adder when you leave proofing switched on. Fourth it divides the combined total by your expected translated-edition sales to get cost per expected sale. Each stage uses only the numbers you entered plus the published bands shown below, so you can repeat the whole calculation on paper and land on the same totals.
Start with the per-word bands the tool uses for planning. Standard commercial translation plans at 8 to 12 cents per word for English to Spanish and English to Portuguese, 9 to 13 cents for English to French and English to Italian, 10 to 14 cents for English to German, and 12 to 18 cents for English to Japanese. Premium literary translation plans at 12 to 18 cents for Spanish, 12 to 17 cents for Portuguese, 13 to 19 cents for French and Italian, 14 to 20 cents for German, and 18 to 25 cents for Japanese. Those bands are deliberately wide because translator seniority and book difficulty genuinely move prices that much, and a single point estimate would pretend at precision the market does not offer. When a vendor quotes inside the band, your budget holds; when a vendor quotes above it, you are paying for seniority, speed, specialization, or extra editing that you should ask to see itemized.
Next the tool multiplies the band by your billable word count. The worked default shows exactly how this behaves. Take 80,000 words into English-to-Spanish standard at 8 to 12 cents: 80,000 times 0.08 equals 6,400 dollars at the low end, and 80,000 times 0.12 equals 9,600 dollars at the high end. That 6,400 to 9,600 dollar interval is the translation-only range before any proofing. Change one input and watch the interval move in straight proportion: every additional 10,000 words adds 800 to 1,200 dollars at this pair and tier, while switching the same 80,000 words to standard Japanese at 12 to 18 cents lifts the translation range to 9,600 to 14,400 dollars without touching anything else. Word count is therefore the dominant lever on total cost, which is why the tool asks for the final manuscript count rather than a draft target or a page count.
Then the tool adds proofreading as a percentage of the translation fee. At standard quality the adder is 15 percent; at premium quality it is 12 percent, reflecting that premium vendors typically bundle more internal revision before the independent pass. Return to the worked default: 15 percent of 6,400 dollars equals 960 dollars of proofing at the low end, and 15 percent of 9,600 dollars equals 1,440 dollars at the high end. Add those to translation and the combined planning total becomes 7,360 dollars at the low end and 11,040 dollars at the high end. If you switch proofing off, the tool sets both proofing figures to zero and the total collapses back to the translation-only range, but you should only do that when a written quote explicitly includes an independent proofing pass by a second linguist. A single translator self-reviewing their own file is not the same control, and unchecking the box to make the total look smaller does not remove the underlying risk.
Finally the tool divides the combined total by expected translated-edition sales. Using the same default with 500 expected sales: 7,360 divided by 500 equals about 14.72 dollars per sale at the low end, and 11,040 divided by 500 equals about 22.08 dollars per sale at the high end. That per-sale interval is the decision figure, because it translates a large one-time production cost into the per-unit burden each future sale must carry. The tool treats roughly 8 dollars per sale at the top of the band as comfortable, 8 to 20 dollars as heavy enough to demand changes, and above 20 dollars as too heavy for the stated sales. Those cutoffs are screening thresholds for planning, not judgments about your book, and they exist so that a small expected audience forces a pause before money moves.
Illustrative hypothetical example 1: a mid-list romance novel into Spanish
Consider an illustrative hypothetical romance novel of 75,000 words, planned from English to Spanish at standard quality, with proofing switched on and 1,200 expected translated-edition sales over two years. The translation range is 75,000 times 8 to 12 cents, which equals 6,000 to 9,000 dollars. Proofing at 15 percent adds 900 to 1,350 dollars. The planning total is therefore 6,900 to 10,350 dollars. Divide by 1,200 sales and the per-sale burden is about 5.75 to 8.63 dollars. The top of that band sits just above the comfortable line, so this project reads as borderline comfortable: workable if the sales estimate is honest, worth trimming if the author suspects 1,200 is optimistic. Practical trims here include confirming the final count after the last developmental pass, since cutting 5,000 words of bloat saves 400 to 600 dollars of translation plus proofing, and asking vendors whether dialogue-heavy romance with consistent terminology can quote toward the middle of the band rather than the top. Nothing about this example is a promise; it is an arithmetic walkthrough showing how word count, band position, and sales interact.
Illustrative hypothetical example 2: a thriller series opener into German at premium quality
Now consider a second illustrative hypothetical case with different pressures: a 95,000-word thriller, English to German, premium quality, proofing on, 900 expected sales. Premium German plans at 14 to 20 cents, so translation is 95,000 times 0.14 equals 13,300 dollars at the low end and 95,000 times 0.20 equals 19,000 dollars at the high end. Proofing at 12 percent adds 1,596 to 2,280 dollars. The planning total is 14,896 to 21,280 dollars. Across 900 sales that is about 16.55 to 23.64 dollars per sale, which lands in heavy-to-too-heavy territory. The instruction from that result is not to abandon German but to change the structure: test standard quality for book one and reserve premium for a retranslation if the series earns it, split the commitment by translating book one only and conditioning book two on sell-through, or build the sales plan past roughly 2,660 copies before premium becomes comfortable at the top of the band. Thriller buyers punish clumsy procedural language, so quality still matters; the point is to buy quality in the right sequence rather than all upfront.
A third hypothetical check: short nonfiction into Portuguese with strong sales
A third illustrative hypothetical rounds out the pattern from the other direction. Take a tight 45,000-word nonfiction title, English to Portuguese, standard quality, proofing on, 2,500 expected sales because the author already speaks to a Brazilian audience. Translation at 8 to 12 cents gives 3,600 to 5,400 dollars. Proofing at 15 percent adds 540 to 810 dollars. The total is 4,140 to 6,210 dollars, and per-sale cost is about 1.66 to 2.48 dollars. That project is comfortable at any point in the band, which frees the author to spend attention on cover adaptation, local testimonials, and category-appropriate pricing instead of agonizing over production cost. Short books with real audiences are the easiest translation approvals an author will ever see, and the math shows why: fewer billable words multiplied by a low per-word band divided across genuine demand.
Reading the language-pair bands like a buyer
Spanish and Portuguese sit at the favorable end of the planning bands because translator supply is deep and commercial workflows are standardized. Expect most qualified vendors to quote inside 8 to 12 cents for standard work, with movement toward the top when your manuscript carries heavy slang, regional wordplay, or dense formatting such as tables and callouts. French and Italian sit one step higher at 9 to 13 cents standard and 13 to 19 cents premium, reflecting slightly thinner literary supply and more involved agreement and register decisions per sentence. German sits higher still at 10 to 14 cents standard and 14 to 20 cents premium, because compound constructions, formal-versus-informal address choices, and long technical sentences demand more editorial judgment per thousand words. Japanese stands apart at 12 to 18 cents standard and 18 to 25 cents premium, driven by script handling, honorific calibration, onomatopoeia choices, and a smaller pool of literary translators working into Japanese from English fiction.
Use those explanations to interrogate quotes rather than to haggle blindly. When a Spanish quote lands at 14 cents with no editing included, ask what justifies the premium over the band top: seniority, a rush schedule, or a genuinely difficult manuscript. When a Japanese quote lands at 13 cents, check what is missing: independent proofing, formatting of vertical text elements, or adaptation of sound-effect language that a low bid quietly excludes. Bands discipline both sides of the conversation. They stop you from accepting an inflated bid out of anxiety, and they stop you from celebrating an underpriced bid that omitted the work you actually need.
Standard versus premium: what the extra money buys
Standard quality in this tool means a clean commercial translation by a qualified translator, followed by a separate proofing pass when you leave that switch on. Sentence meaning carries over accurately, dialogue reads naturally, grammar and spelling are correct, and formatting survives the transfer. For most plot-driven commercial fiction and most practical nonfiction, standard with independent proofing is sufficient, provided the translator works in their native language and has published credits in your category. Premium means a styled literary translation with closer attention to voice, rhythm, register shifts, humor timing, and cultural adaptation, plus heavier internal revision before proofing. The per-word premium of roughly 50 percent reflects that second layer of editorial attention, not just a more famous name on the invoice.
Choose premium when voice is the product. Literary fiction, memoir with distinctive narration, humor-driven romance, and prestige fantasy with invented registers lose disproportionate value in flat translation, so the premium band is usually justified if the sales support it. Choose standard when plot, information, or series momentum is the product. Fast-paced thrillers, cozy mysteries with formulaic structures, practical business books, and mid-list romance with strong dialogue engines typically perform well in standard with rigorous proofing, especially when you reinvest the savings into cover adaptation and local launch promotion. Never choose premium as a substitute for proofing: even the most gifted literary translator benefits from a second set of eyes, which is why the tool keeps a 12 percent proofing adder at premium rather than zeroing it.
Why the proofing adder stays on by default
The proofing switch defaults on because skipping independent review is the most common false economy in author translation budgets. Translation creates errors at a predictable rate: homophone slips, dropped negatives, misread idioms, inconsistent character names, timeline contradictions introduced mid-file, and formatting damage around italics, scene breaks, and chapter headings. A translator self-review catches many of these but systematically misses the ones shaped by their own misreading, which is exactly what a second linguist is positioned to find. Budgeting 15 percent at standard and 12 percent at premium reflects market practice for a clean manuscript; heavily formatted books or manuscripts with tracked-change residue can push real proofing quotes above those percentages, which is another reason to treat the tool output as a floor for planning rather than a ceiling.
Turn proofing off in the tool only when a vendor contract explicitly names an independent proofreader or a separate editing stage performed by someone other than the translator. A line in a quote that says edited delivery without naming the second person is not sufficient evidence. Ask for the proofreader credit, the round count, and whether proofing covers comparison against the English source or target-language polish only. Comparison proofing costs more and catches meaning drift; target-only polishing reads smoothly while potentially preserving mistranslations. Your budget should reflect the control you actually need, and for a first translation into a new language, comparison proofing by a second native speaker is the control most authors actually need.
Cost per expected sale: the figure that approves or stops the project
Total cost tells you what to budget; cost per expected sale tells you whether to proceed. A translation is a fixed upfront investment repaid one royalty at a time, so dividing the planning total by honest expected sales converts an intimidating lump sum into a per-unit hurdle you can compare against realistic per-copy earnings. At the default 7,360 to 11,040 dollar total across 500 sales, the 14.72 to 22.08 dollar per-sale burden exceeds what most single-copy royalties can repay within two years, which is why the tool flags it as too heavy. The same total across 2,000 sales falls to roughly 3.68 to 5.52 dollars per sale, a burden most translated editions can carry. Volume is the great normalizer of fixed production cost, and no choice of vendor changes that arithmetic.
Set the sales input with discipline. Count only the translated edition in the target language, not your English backlist, not audio, and not box sets in other languages. Use a two-year window because translation royalties arrive slowly through retailer reports and rights-holder statements, and a ten-year fantasy horizon can justify almost any spend on paper. Ground the number in comparable evidence: newsletter subscribers in the target language, prior translated-title performance, preorder or waitlist signups from that audience, and any retail preorders actually recorded. If you cannot name the readers who will produce the first few hundred sales, enter a lower number and observe how quickly the verdict deteriorates. That deterioration is information, not pessimism, and acting on it early is far cheaper than discovering it after paying for 90,000 words.
Edge cases that break naive estimates
Very short books expose minimum fees. Many translators and agencies enforce a project minimum between 300 and 800 dollars regardless of word count, so a 3,000-word picture book or a 8,000-word lead magnet can quote above the per-word band math. Enter the real count, note the tool total, then ask vendors for their minimum before treating the band as final. Very long books expose schedule surcharges instead: a 180,000-word epic cannot be translated carefully in three weeks by one person, so rush fees, team-splitting costs, and terminology-management overhead push real quotes toward or above the band top. Plan twelve to sixteen weeks for a full-length novel and longer for Japanese, and get schedule terms in writing alongside the per-word rate.
Repetitive series content creates a legitimate discount case the tool does not auto-apply. Books two and three that reuse invented vocabulary, recurring characters, and established style guides genuinely cost less per word to translate well, because glossary and memory assets carry forward. Raise this explicitly with vendors rather than assuming it: ask for a translation-memory discount schedule, a glossary reuse credit, and a declining per-word rate across contracted volumes. Conversely, books with heavy formatting, footnotes, endnotes, indexes, recipes, exercises, or illustrated callouts cost more than clean narrative at the same word count, because desktop work and cross-checking consume hours that per-word math ignores. Disclose the file exactly as it is, including images and tables, when requesting quotes.
Machine-translation post-editing deserves a clear warning. Some vendors offer steep discounts for editing machine output rather than translating from scratch, and for highly repetitive technical text that workflow can be legitimate. For voice-driven fiction, humor, memoir, and marketing-sensitive nonfiction, post-edited machine output routinely needs so much rewriting that the final cost approaches human translation while the quality trails it. Do not enter a machine-editing rate into this tool and compare it against human bands; the outputs are not the same product. If you test post-editing at all, do it on a paid 2,000-word sample judged blind by a native reader in your category, and only then discuss full-manuscript terms.
Rights complications can strand a paid translation entirely. Translating a book you do not control in the target language, or translating before confirming that your distribution path accepts the language edition, converts production spend into an unsellable file. Confirm that your publishing agreement, any co-author agreement, and any prior rights grant leave the target language available, and confirm that your distributor or retailer path supports the language, the trim requirements for print, and any tax or pricing constraints in that market. These checks cost nothing and belong before any vendor deposit. A translation budget is only the second gate; rights clearance is the first.
Failure modes authors repeat
The most expensive failure mode is approving a translation on English-edition enthusiasm rather than target-language evidence. An English title earning steady royalties does not automatically earn readers in Berlin or São Paulo, where covers, categories, price expectations, and competitive titles all differ. The tool cannot see that gap, which is why the sales input exists: it forces you to state the demand thesis as a number and watch the per-sale burden respond. A second failure mode is buying premium quality for a book whose sales cannot carry standard, usually out of perfectionism about a beloved manuscript. Craft pride is admirable and budget-agnostic math is not; translate the lead title at standard, prove the audience, then fund premium for the reissue or the sequel.
A third failure mode is unbundling proofing to hit a round-number budget. Saving 960 to 1,440 dollars on an 80,000-word Spanish project feels tangible on invoice day and costs far more in returns, one-star reviews in the target language, and retailer sample damage that suppresses conversion for months. Target-language readers cannot see your English reviews; they see the sample of the book you actually published in their language. A fourth failure mode is translating the whole series before book one earns anything. Series translation multiplies every word-count error by the number of volumes, so condition later volumes on measured sell-through of the first translated book plus genuine audience signals such as newsletter growth and review velocity in that language. Staging is not timidity; it is how fixed-cost series math stays solvent.
A fifth failure mode is treating the planning band as a quote and skipping competitive bids. Bands describe the market middle for budgeting; individual vendors quote above or below for defensible reasons including seniority, specialization, schedule, and included editing. Collect two or three written quotes that each specify the per-word rate, the billable count methodology, the number of revision rounds, the named proofing stage, the delivery schedule, and the payment milestones. When quotes cluster inside the tool band, proceed with confidence. When they cluster above it, either the manuscript is harder than you assumed or the schedule is too tight. When one bid sits far below the rest, investigate scope gaps before celebrating savings.
Genre-specific guidance
Romance translates at high volume and rewards consistency over flourish. Standard quality with rigorous proofing carries most contemporary, historical, and paranormal romance well, provided the translator has category credits and your style sheet pins down terms of endearment, heat-level vocabulary, and recurring location names. Watch word-count creep from steamy dialogue tags and repeated intimate scenes: 5,000 untrimmed words cost 400 to 600 dollars of Spanish standard translation before proofing, so a final tightening pass in English pays for itself twice, once in English pacing and once in avoided translation spend. For series romance, negotiate glossary reuse across books and translate the highest-converting entry point first rather than volume one by default when a later book clearly leads the funnel.
Mystery, thriller, and crime fiction punish procedural errors more than any other commercial category. Weapon handling, forensic vocabulary, rank structures, legal procedure, and technology references must land exactly, because genre readers detect mistakes instantly and review them harshly. Prefer translators with demonstrated crime credits even at standard tier, supply a terminology list for every recurring procedural term, and keep proofing switched on with source comparison. Do not economize by stripping proofing from a 90,000-word police procedural to save 12 to 15 percent; a single mistranslated forensics passage can do more reputational damage in the target market than the proofing fee ever cost.
Fantasy and science fiction carry two surcharges the per-word band understates: invented-language management and world-consistency labor. Constructed terms, magic systems, starship schematics, and map-label consistency require glossary construction, cross-volume tracking, and consultation rounds that straight narrative does not. Budget toward the top of the band, add explicit glossary hours to vendor scopes, and plan longer schedules, particularly for Japanese where transliteration choices for invented names ramify across covers, interiors, and marketing copy. Translate book one only until the audience proves itself, and freeze the glossary before contracting book two so later volumes inherit decisions instead of relitigating them at full price.
Nonfiction trades word-count efficiency for formatting and verification costs. A tight 45,000-word business or self-development book can be the cheapest comfortable approval in this tool, yet real invoices rise when manuscripts include tables, charts, worksheets, checklists, citations, and index entries that each need handling. Supply clean source files with editable tables rather than flattened images, confirm whether index and footnote localization are included, and verify that exercises and frameworks read naturally rather than literally in the target language. Testimonial and case-study adaptation deserves explicit scope: replacing unrecognizable English examples with locally resonant ones is editorial work beyond translation, and it should be quoted as such.
Memoir, literary fiction, and poetry-adjacent prose justify premium most often, because voice constitutes the buying reason. Budget premium bands, insist on sample translations of the same 1,500-word passage by two shortlisted translators, and judge those samples blind with a native reader who understands the category. Brief translators on register, era, class markers, and humor mechanics rather than assuming they will infer them; a memoir narrated in working-class vernacular fails when rendered in neutral formal prose, and no proofing pass restores a flattened voice. Keep expected sales brutally honest here, because literary translations earn slowly and the per-sale math is unforgiving at low volume.
Children's books invert the usual word-count logic. A 1,200-word picture book looks trivial in per-word math yet demands read-aloud rhythm, rhyme or rhythm compensation, sound-effect adaptation, and text-to-illustration fit that experienced children's translators charge minimums to provide. Expect quotes above the band-derived total, budget for read-aloud testing with native children, and confirm who adapts cover and interior lettering. Early-reader and middle-grade series behave more like commercial fiction per word but still need age-band vocabulary calibration, so favor translators with verifiable children's credits over generalists quoting the band midpoint.
From estimate to decision: pairing cost with market return
This calculator answers the production question: what will the words cost to convert. It deliberately does not answer the market question: what will the converted book earn in a specific territory. Approving a translation requires both, because a cheap production cost in a market with no demand still loses money, and an expensive production cost against proven demand can still be the right investment. The honest sequence is to budget here first, then carry the planning total into a rights-and-royalty comparison that weighs likely royalties, retailer cuts, withholding, and print or audio adaptation against that fixed cost. Stating the pairing explicitly keeps each tool honest: per-word estimation here, market return judgment there, neither pretending to do the other's job. These planning bands describe production outlay for budgeting, not quotations and not forecasts of market performance, and any vendor or advisor who conflates the two is selling certainty the data cannot support.
In practice that means three decision paths. When both the per-sale burden here is comfortable and the market comparison supports the royalty case, approve and move to vendor sampling. When the per-sale burden is comfortable but the market case is thin, pause and build demand evidence first: audience growth, preorder signals, or a lower-cost test such as a translated novella or sample. When the per-sale burden is heavy or worse, do not rescue the project with optimistic royalty assumptions; restructure the production side by trimming words, stepping down a quality tier, choosing a cheaper pair with proven audience overlap, or staging one book instead of the series. Cost discipline belongs before market optimism, because production invoices arrive with certainty while royalties arrive with variance.
Budgeting workflow that survives contact with vendors
Run the tool, record the low-to-high total, and set your internal budget at the top of the band plus a 10 percent contingency for scope discoveries such as hidden formatting, index work, or glossary construction. Request quotes using a fixed brief: final word count with counting methodology, target language and regional variant, quality tier with sample expectations, explicit independent proofing, file formats and formatting scope, glossary deliverables for series, schedule with milestones, revision-round count, and payment terms. Score samples blind with a native category reader before discussing price, because a cheap bid on a failed sample is not actually cheap. Only after samples rank do you compare numbers, and only quotes that specify every brief element qualify for comparison.
Negotiate structure, not just rate. A vendor who refuses to move on per-word price may still concede a second revision round, a named proofreader, glossary delivery in reusable format, or a series discount schedule, each of which lowers lifetime cost more than a half-cent rate cut. Lock the billable count methodology in the contract: which file counts, how repetitions and translation-memory matches are treated, and how author-side additions during production are priced. Price every change path before it happens, because mid-project additions at rush rates are where translation budgets most often rupture. Keep a written log of scope decisions so the final invoice reconciles against the quoted band rather than against memory.
Frequently asked depth: the questions behind the questions
Authors rarely ask only what translation costs; they ask whether their specific book should be translated now, into which language first, and at what quality. Take languages first. The cheapest pair is not automatically the best first move; audience overlap decides. A romance author with an existing Spanish-language newsletter should favor Spanish even if Portuguese bands match on price, while a thriller author with German also-boughts and newsletter traction should favor German despite the higher band. Let evidence rank languages, then let this tool price the ranking. Translating into a language where you have no reader pathway converts a production bargain into a distribution problem, and distribution problems cost more than per-word savings ever recover.
Timing questions follow the same evidence-first logic. Translating at launch multiplies unknowns: unproven English performance, unproven cover, unproven positioning, all carried into a second language at full fixed cost. Translating after the English edition proves conversion, accumulates reviews, and clarifies the best-performing entry point de-risks the fixed spend with information you already paid to acquire. Series authors should translate the book that sells the series, which is usually book one but occasionally a prequel novella or a breakout later volume; the tool prices whichever manuscript you enter, so enter the true funnel leader rather than defaulting to volume one out of habit.
Quality anxiety drives many questions, and the answer is comparative sampling rather than abstract reassurance. Shortlist two translators, pay both for the same 1,500-word sample covering dialogue, exposition, and one difficult passage, and have a native category reader rank the samples without seeing prices. When samples tie, buy the cheaper bid with stronger schedule and proofing terms. When the premium sample clearly preserves voice the standard sample flattens, check whether the per-sale math supports the upgrade at honest sales; if it does not, translate at standard now and earmark a premium retranslation as a stretch goal funded by actual translated-edition earnings. Quality ladders beat quality gambles.
Rights, tax, and pricing questions sit adjacent to cost and deserve direct handling without overreach. Confirm language availability in every agreement touching the book before spending, verify that your distribution path supports the target language in both digital and print where relevant, and research local price expectations so your royalty-per-copy assumption reflects the market rather than a currency conversion of your English price. This page offers no tax or legal advice and no market forecasts; it gives you the production arithmetic to carry into qualified professional advice and real market data. Bring the planning total, the per-sale burden, and your evidence-backed sales estimate to those conversations, and you will get better answers than any generic forecast could provide.
Length questions close the loop. Authors often ask whether to trim before translating, and the math answers plainly: every 1,000 words removed saves roughly 80 to 120 dollars of Spanish standard translation plus proofing, scaling higher for harder pairs and premium tiers. A disciplined final edit that removes 8,000 words of repetition saves around 736 to 1,104 dollars at Spanish standard with proofing, which often exceeds the cost of the editing pass itself. Cut in English where the story allows, freeze the manuscript, then count and translate the frozen file. Translating a moving manuscript and paying for revised words twice is a budgeting wound that never needed to happen.
Where the work continues
When the estimate looks workable and the per-sale burden fits your honest sales outlook, continue the work in Author Desk at /author-desk, where manuscript preparation, vendor coordination, and edition planning stay organized in one place. Use that workspace to freeze the final word count, store competing vendor quotes against the planning band from this page, and track the proofing and formatting milestones that turn a translated file into a publishable edition. Cost estimation ends the budgeting phase; the production phase that follows deserves the same discipline, and keeping quotes, glossaries, and schedules together is how that discipline holds.