A Logfile is wxrks' word, character, and segment count and analysis for a project. It's the basis for project quotes, vendor budgets, and understanding how much of a job is genuinely new work versus content your Translation Memory can already cover. This article explains what a Logfile shows, how to read its breakdown, and how to download or create one.
💡 Who is this for? This guide is for Account Admins and Project Managers who need to read a project's Logfiles to understand word counts, TM match rates, and cost estimates before quoting a project or paying vendors within the wxrks platform.
Where to find a project's Logfiles
Open the project and click its Logfiles tab. Every Logfile ever generated for that project is listed here, with Name, Type (Pre-analysis or Post-analysis - see below), Creation Date, Targets (target language), and Workflow columns.
A Logfile used for project costs badge under a Logfile's name means it's the Logfile currently used for the project's costs - the one feeding the Cost Results card on the project's Costs tab. Each row's Status shows whether the count finished processing. Once it shows Completed, the row offers Summary, which opens the Logfile's totals by language (see Reading a Logfile's summary below), and Per-file analysis, which breaks the same count down by file (see See the analysis of each file below). The ⋯ menu at the end of the row holds Download, Rename and the other actions for that Logfile.
Above the list, three toggles change how repetitions are counted in every Logfile's breakdown: TM Match Over Repetition (when on, 100%/101% Translation Memory matches are no longer counted as repetitions - each one keeps its own match-tier count instead), Non-Translatable Match Over Repetition (when on, every non-translatable segment counts as non-translatable, even if it repeats elsewhere - it's never folded into the repetition count), and Confidence Score Match Over Repetition (applies to segments wxrks approved automatically because their confidence score met your minimum after pre-translation: when on, they're counted as Confirmed by Confidence Score even when they're repetitions; when off, repetitions keep their own count and only the remaining approved segments are counted as Confirmed by Confidence Score).
Pre-analysis vs. post-analysis
A Logfile's Type is either:
Pre-analysis: generated before translation starts, on the original source content. It's what a quote is based on - the word/character counts a Project Manager or Quote Automation uses to price the job.
Post-analysis: generated after the linguist or reviewer has done the work, comparing the result back against the pre-analysis Logfile. It shows the actual effort spent, which is what vendor payables are calculated from.
Reading a Logfile's summary
On the Logfile's row, click Summary.
The Words counted from badge, for example Words counted from Original (source), shows whether this Logfile counted words from the original source content or from pre-translated target content - see Calculating Costs and Vendor Payments from Target Content for why you'd want the Target option and how it changes cost calculation.
Next to it, Net rate names the coefficient set the project uses. The Workflow selector picks the workflow step whose Net Rate Coefficients weight the numbers, using the project's current coefficients. Below it, each target language gets its own row with its Segments, Words, Net words and Characters, and an arrow at the start of the row that expands the full match-tier breakdown.
Words is the total word count. Net words is that count weighted by the coefficients of the selected workflow step, the same value as the Net of the Total row in the breakdown. Switch Workflow to see the net words of another step. For example, if 400 of a language's 1,000 words are 100% matches and the other 600 are new, and the Review step's coefficients are 20% for 100% matches and 100% for new words, the row shows 1,000 Words and 680 Net words for Review (600 × 100% + 400 × 20%). After a Sync net rate (see below), Net words is where the new coefficients show.
Choose what the tables show
When you open the details of a language, a Show bar sits above the table: Segments, Words, Characters and %. Turn off the counts you don't need to make the table easier to read. Your choice is remembered in your browser for the next Logfile you open, and at least one option always stays on.
The % column shows each row's share of the Total. Its header says what it is based on: % of Words when the analysis counted words, otherwise % of Characters, and % of Segments when there are neither words nor characters. In the example below, 51.9% of the words have no useful Translation Memory match (0% - 49%), and 33.3% are 101% matches.
If you are a Vendor, open your task and go to its Logfiles tab: the same Show bar and % column are there, so you can see at a glance how much of the task is new content and how much is covered by repetitions or Translation Memory matches before you start. Vendors can read Logfiles but can't rename them.
See the analysis of each file
The summary adds up every file of the project. To see how much of each file is new content, repetitions or Translation Memory matches, for example to find the file that drives most of the cost:
On the project's Logfiles tab, click Per-file analysis on a Logfile whose Status is Completed.
The window lists one row per file and target language, with its Words. Click a file to expand its match-tier breakdown, with the same Show bar and % column as the summary.
ℹ️ Note: The per-file breakdown shows raw counts only, with no Net column or Net Rate Factor: net values are calculated per workflow step for each language, so they appear only in the Summary.
Rename a Logfile
Give a Logfile a name that says what it was for, such as Initial quote or After source update:
On the project's Logfiles tab, click ⋯ on the Logfile's row and choose Rename.
Type the new name and click Save. The row then shows who renamed it, when, and the previous name, for example Renamed by Jane Doe · 2026-10-26 17:32 · was "Project Logfile 1".
Renaming requires the Project Update permission; Vendors don't see the option. Save stays disabled while the name is the same as the current one.
Sync the project's Net Rate coefficients
A Logfile's Net words come from Net Rate Coefficients: the effort percentage applied to each match tier, for each workflow step. The first time a project needs them, wxrks saves a copy of the coefficients that apply to it: the coefficient set selected for its Organizational Unit (the client or department the project belongs to), otherwise your account's default set, otherwise the wxrks built-in default. Later changes to those coefficients don't reach the project, so its counts don't shift while the work is under way.
When the project should use the current coefficients, for example after you corrected a percentage on the Organizational Unit's set, sync them from the Logfiles tab:
On the project's Logfiles tab, click Sync net rate, next to New Logfile.
Review the Sync net rate coefficients window before you confirm:
Now and After sync name the coefficient set the project uses today and the one it will use.
The count above the table says how many coefficients change. The table lists them by Workflow and range, with the TM, MT and NT columns. A changed value shows as before → after, for example 60 → 50.
Turn on Show unchanged coefficients to see the full table.
Click Sync coefficients. A message confirms the sync and reminds you to use Recalculate Budget on the tasks.
When the project already uses the current coefficients, the window says The project already uses the current coefficients of the organizational unit. and Sync coefficients stays disabled.
What the sync changes, and what it doesn't:
Logfiles: every Logfile of the project, including the existing ones, shows its Net words with the new coefficients right away, in its Summary. Its match ranges stay the ones of the analysis that created it, so if the new set has different ranges, create a new Logfile (see Creating a Logfile manually below).
Project costs: they don't change.
Task budgets: they change only when you run Recalculate Budget on a task.
For example, if the Organizational Unit's set lowers the TM coefficient of the Translation step's 50% – 99% range from 60 to 50, as in the window above, a target language with 1,000 words in that range shows 600 net words before the sync and 500 after it. Task budgets keep the previous coefficients until you recalculate them.
Sync net rate requires the Recalculate Task Budget permission, which Account Admins and Project Managers have by default. Clients and Vendors don't see the button.
The match-tier breakdown
In the Summary, click the arrow at the start of a language row to expand its full breakdown.
Segments, words, and characters are each broken down by:
Total: everything in the analyzed file(s).
Confirmed by Confidence Score: segments wxrks approved automatically because their confidence score met your minimum. This row sits directly under Total, above Non-translatable and Repetition, and has its own Net Rate Factor. It appears only when Include in the cost analysis was on when the analysis ran; Logfiles created before the feature was enabled keep their original layout. See Auto-Confirming Segments by Confidence Score.
Repetition: content that repeats elsewhere in the same job (unless one of the toggles above changes that).
Non-translatable: content excluded from translation entirely.
101% / 100% / fuzzy-match ranges (for example 95%-99% down to 0%-49%): Translation Memory match tiers - how closely each remaining segment matches something already in your TM, from an exact context match (101%) down to no useful match at all (the lowest range). The fuzzy-match ranges come from the Net Rate Coefficients in use.
Every count has two columns: All (the raw count) and Net (that count after applying the project's Net Rate Coefficients for the selected Workflow - the per-match-tier pricing weight used for cost and payable calculations). The Net Rate Factor column on the right shows that weight itself for each tier - for example, if your Organizational Unit discounts 100% TM matches to 20% of a full word's price, a segment in the 100% tier shows a Net Rate Factor of 0.20, and its Net count is 20% of its All count. Configure these coefficients from your Organizational Unit's Net Rate tab - see All about Organizational Units.
Downloading a Logfile
On the project's Logfiles tab, click ⋯ on the Logfile's row and choose Download.
Under Output, choose how to split the counts:
One file with all languages (default): a single file with the counts of every target language.
Split by target language (ZIP): one file per target language, packaged in a .ZIP. Use it when each language is quoted, invoiced or imported into another system separately.
Under Format, choose CSV, XLSX or LOG. LOG is the format XTRF imports; see below.
Click Generate. A progress bar shows while the files are generated, and the download starts when they are ready.
Split files are named <project code>_<source>_<target>.<format>, for example B-26295-77831-CCA2_en-US_pt-BR.csv. A target language with no content in the analysis gets no file, and the window lists the languages left out.
LOG format for XTRF
LOG produces the analysis in the layout of the Phrase (formerly Memsource) analysis log, the one XTRF reads to calculate TM savings. Use it when your team prices projects in XTRF and needs one analysis per target language.
LOG always comes out as one file per target language: with one target language you get the .log file itself; with more than one, a .ZIP with one .log per language.
Each .log file has one block per project file and a final Total block. The Language direction line uses Phrase's language codes, for example en_us > pt_br. The wxrks match tiers fill the LOG lines like this:
LOG line | What it counts from the wxrks analysis |
Translated | Non-translatable segments |
Context Match | TM 101% matches |
Repetitions | Repetitions |
Format Change | Always 0 (wxrks does not track this) |
100%, 95% - 99%, 85% - 94%, 75% - 84%, 50% - 74% | The same tiers |
No Match | TM 0% - 49% and new content |
In a post-analysis, TM, MT, context-sensitive translation and non-translatable counts of the same tier add up on one line, and segments confirmed by confidence score count as 100%. If your account uses custom match tiers, each one goes into the LOG line that contains its lower limit.
💡 Tip: To import the file in XTRF, upload it in the receivable or payable's TM Savings and choose Memsource as the CAT tool.
Creating a Logfile manually
wxrks generates pre-analysis and post-analysis Logfiles automatically at the right points in a project's workflow, but you can also create one on demand - useful for re-checking word counts after a Translation Memory update, or for generating a Target Logfile once pre-translation is done. On the project's Logfiles tab, click New Logfile, next to Sync net rate. The New Logfile window creates a pre-analysis of every file of the project:
Logfile name: leave it empty and wxrks names it "Project Logfile {number}".
Ignore Confirmed Segments and Ignore Locked Segments: exclude segments already confirmed or locked in the Editor from the count.
Non-Translatable Match Over Repetition, TM Match Over Repetition, and Confidence Score Match Over Repetition: the same repetition-counting behavior as the three toggles on the Logfiles tab, set here for this one Logfile instead of the whole project.
Count words from - Original (source) or Translated (target): whether this Logfile counts the original source content or pre-translated target content. See Calculating Costs and Vendor Payments from Target Content for when to use the Target option and how it's used for cost calculation.
Click Create.
Related articles










