Job Description Parsing API Comparison 2026: 8 APIs

A job description parsing API turns a job ad into structured data. Some APIs return the fields of the posting (title, location, salary, skills, requirements): Textkernel, RChilli, HrFlow.ai, DaXtra, Affinda, Hirize and JANZZ. Others return the hiring criteria you need to score candidates: HireLayer Job Extract returns weighted, explained criteria ready for matching. Pick the first to publish or index jobs, the second to screen applicants.

HireLayer is our product. The facts about other vendors come from their own docs and pricing pages, checked on 7 October 2026 and linked at the end.

The short list

API	Returns	Languages	Published price
HireLayer Job Extract	Weighted criteria with a mandatory flag and a rationale	Any input language; labels in French	1 credit, from €0.0445; 50 free a month
Textkernel Job Parser	Structured job, optional title and skills normalization	9	1 credit per parse; plans from $99 a month
RChilli JD Parser	75+ fields, from a file or text	26	1 credit per parse; 500 credits for $75 a month
HrFlow.ai text parsing	Job object: skills, tasks, requirements, ranges, benefits	43+ (general claim)	Not published
Affinda Job Description Parser	Structured job description	Not published	Not published, through sales
DaXtra job orders	Structured XML or JSON	40+ (data extraction)	Not published, through sales
Hirize Job Parser	Job data with hard, soft and enriched skills	24+	1 credit per file; credit price in a calculator
JANZZparser!	50+ categories incl. occupation, skills, salary	60+	Licence on request

Fields or criteria: two kinds of output

A field parser reads a job ad the way a job board needs it: title, company, location, contract type, salary range, required skills, education and experience. That is what you want to fill a job form automatically, index postings for search, or feed analytics.

A criteria extractor reads the same ad the way a recruiter screens candidates: what must a candidate have, what is a plus, and how important is each point. The output is a short list of criteria with weights, ready to be checked against each resume. It is less complete as a record of the job, but far more usefull for matching.

The APIs in detail

HireLayer Job Extract

Job Extract takes the job description as text, up to 50,000 characters in any language, and returns criteria. Each criterion has an id, a short label, a weight from 1 (nice to have) to 3 (essential), a flag saying whether the ad states it as a hard requirement, and a rationale. The list goes straight into HireLayer Match to score a resume criterion by criterion. One call costs one credit. Labels and rationales are written in French in the current version.

Best for: screening and ranking applicants against a job. It is not built to fill every field of a job posting.

Textkernel Job Parser

Textkernel's Tx Platform parses job orders into a structured job in 9 languages, for 1 credit per parse. Add-ons normalize the job title (+0.2 credit) and skills (+0.1), and OCR a scanned job (+0.05). Plans start at $99 a month and the trial gives 500 free credits.

RChilli JD Parser

RChilli parses a job file (ParseJD) or plain text (ParseJDText) into more than 75 fields, in 26 languages. Each parse costs 1 credit, plus 1 when a file needs OCR. Credits start at $75 a month for 500, with 100 free credits to start.

HrFlow.ai

HrFlow.ai parses raw text into a job object with name, location, summary, responsibilities, requirements, skills, languages, certifications, ranges, benefits, culture and tasks. The job can then be indexed in a Board for HrFlow's scoring. Prices are not published.

Affinda, DaXtra, Hirize, JANZZ and Lightcast

Affinda documents a job description endpoint in its platform API, but sells it with the resume parser through its sales team, and its pricing page excludes it. DaXtra parses job orders through the same REST service as resumes; access goes through a demo. Hirize charges 1 credit per job file and lists 24+ languages. JANZZ.technology's JANZZparser! covers job ads and CVs in more than 60 languages under a licence. Finaly, Lightcast classifies job postings in English only and extracts skills from posting text, under a commercial licence.

How to choose

Decide what the output is for. Publishing and search need fields; screening needs criteria with weights.
Test on your own ads. Job descriptions mix company marketing, legal text and the real requirements. Check that the parser keeps the requirments and drops the rest.
Check normalization. A skill or title only helps search and matching once it maps to a stable id; see our skills taxonomy API comparison.
Count the full flow. If the parsed job feeds matching, compare the cost of parsing plus matching, not parsing alone. Our candidate matching API comparison covers the second step.

To try criteria extraction on one of your job ads, the HireLayer Job Extract page shows a full request and response.

Try the live demo Read the API documentation View pricing

Frequently asked questions

What is a job description parser?

Software that reads a job ad and returns structured data: either the fields of the posting, such as title, location and required skills, or the hiring criteria used to evaluate candidates.

Which job description parsing APIs publish their prices?

HireLayer (1 credit per call, from €0.0445), Textkernel (1 credit per parse, plans from $99 a month) and RChilli (1 credit per parse, 500 credits for $75 a month). Affinda, HrFlow.ai, DaXtra and JANZZ sell on request.

Can a job description parser help match candidates?

Yes, if it returns criteria or normalized skills. HireLayer Job Extract returns weighted criteria that HireLayer Match checks against each resume; field parsers usually feed an index that a separate matching engine searches.

Does job description parsing work in other languages?

Support varies: Textkernel lists 9 languages for job parsing, RChilli 26, Hirize 24+ and JANZZ 60+. HireLayer accepts job descriptions in any language but writes its criteria in French. Lightcast's posting classification is English only.

Sources and further reading

Louis Desclous

Published on October 7, 2026 · Reading time: 7 minutes