Most pages about the SBI Clerk cut-off print a table of state-wise marks. This one does not, and the reason is simple: we could not find any document on SBI's website that states cut-off marks, and our rule is to print only figures we have read in an official source. What we can give you accurately is how the cut-off is formed, what SBI does publish, and the full state-wise vacancy table for the newest cycle, which is the single biggest input into any cut-off.
The newest notified cycle as of 13 November 2026 is Advertisement No. CRPD/CR/2026-27/17, dated 11 August 2026, for Junior Associates (Customer Support & Sales).
What SBI publishes, and what it does not
We read the SBI careers page on 13 November 2026 and opened the result links for the last three Junior Associates recruitments.
| Item | What we found |
|---|---|
| Prelims result | A page titled "Marks Secured by the Candidate (Phase-I)". It asks for a roll or registration number, a date of birth and a captcha. It shows one candidate's own result. |
| Mains / final result | A PDF headed "Final Result - Roll Numbers of Provisionally Selected Candidates for Appointment". The 2025-26 file we opened is a list of roll numbers only. |
| Marks after the final result | Another individual lookup, "Marks Secured by the Candidate". |
| Wait lists | Further roll-number lists, with their own individual marks lookups. |
| A public cut-off list by state and category | Not found for 2026-27, 2025-26 or 2024-25. |
So the position is this. A candidate can see their own marks after logging in. A public, openable document with cut-offs for every state and category is not on the careers page. Figures that circulate elsewhere may be accurate, but they come from candidates' own score cards, and we cannot verify them against an SBI document. We reproduce none.
How the Prelims qualifying line is formed
The advertisement describes the Prelims shortlist in three sentences, which we summarise:
- There are no minimum qualifying marks for any individual test or for the aggregate, and section-wise marks are not maintained.
- All candidates are arranged in descending order of aggregate marks out of 100.
- In each category, approximately 10 times the number of vacancies, subject to availability, are shortlisted for Mains.
Because you apply for one State or UT and the merit is state-wise and category-wise, the line falls at a different mark in each state and each category. The "cut-off" is simply the aggregate of the last candidate shortlisted there. It is an outcome, not a number fixed in advance.
Arithmetic, not an SBI figure: for a state with 104 General vacancies, as Uttar Pradesh has in 2026, ten times is about 1,040 General-list places in Mains. For 608 General vacancies, as in Tamil Nadu, it is about 6,080. The advertisement says "approximately" and "subject to availability", so the real numbers can differ.
How the final line is formed
- Prelims marks are not added. Only the Mains aggregate, out of 200, counts.
- You must reach a minimum percentage on the Mains aggregate that the Bank decides. SC, ST, OBC, PwBD, ex-servicemen and disabled ex-servicemen candidates get a 5% relaxation. The advertisement does not print the percentage.
- The merit list is drawn state-wise and category-wise in descending order of Mains aggregate, up to the number of vacancies.
- Provisional selection is subject to the local language test, where it applies. Failing it means no appointment, however high the marks.
- A wait list of up to 50% of vacancies, state-category wise, is kept for one year and released quarterly against non-joining and resignations. This is why SBI publishes first, second and third wait lists after the main result: the last selected mark moves down a little with each list.
Vacancies in the 2026 cycle (Advertisement 17)
Regular vacancies, from the advertisement's table. SBI states that these are provisional and may vary with the Bank's actual requirement.
| State / UT | Local language(s) to opt from | SC | ST | OBC | EWS | General | Total |
|---|---|---|---|---|---|---|---|
| Gujarat | Gujarati | 23 | 49 | 89 | 33 | 136 | 330 |
| Karnataka | Kannada | 121 | 53 | 205 | 76 | 305 | 760 |
| Madhya Pradesh | Hindi | 16 | 22 | 16 | 11 | 48 | 113 |
| Chhattisgarh | Hindi | 34 | 91 | 17 | 28 | 117 | 287 |
| Odisha | Odia | 176 | 242 | 132 | 110 | 440 | 1,100 |
| Jammu & Kashmir | Urdu / Hindi | 6 | 9 | 22 | 8 | 39 | 84 |
| Himachal Pradesh | Hindi | 2 | 0 | 1 | 0 | 6 | 9 |
| Ladakh | Urdu / Ladakhi / Bhoti (Bodhi) | 0 | 0 | 1 | 0 | 6 | 7 |
| Tamil Nadu | Tamil | 267 | 14 | 380 | 141 | 608 | 1,410 |
| Telangana | Telugu / Urdu | 41 | 18 | 70 | 26 | 105 | 260 |
| Rajasthan | Hindi | 59 | 45 | 70 | 35 | 141 | 350 |
| West Bengal | Bengali / Nepali | 127 | 27 | 122 | 55 | 224 | 555 |
| Andaman & Nicobar Islands | Hindi / English | 0 | 1 | 4 | 1 | 9 | 15 |
| Sikkim | Nepali / English | 3 | 12 | 14 | 6 | 25 | 60 |
| Uttar Pradesh | Hindi / Urdu | 52 | 2 | 67 | 25 | 104 | 250 |
| Maharashtra | Marathi | 171 | 154 | 463 | 171 | 759 | 1,718 |
| Goa | Konkani | 1 | 7 | 11 | 6 | 37 | 62 |
| Arunachal Pradesh | English | 0 | 24 | 0 | 5 | 26 | 55 |
| Assam | Assamese / Bengali / Bodo | 7 | 13 | 29 | 11 | 50 | 110 |
| Manipur | Manipuri / English | 0 | 4 | 1 | 1 | 8 | 14 |
| Meghalaya | English / Garo / Khasi | 0 | 10 | 1 | 2 | 11 | 24 |
| Mizoram | Mizo | 0 | 14 | 1 | 3 | 14 | 32 |
| Nagaland | English | 0 | 12 | 0 | 2 | 13 | 27 |
| Tripura | Bengali / Kokborok | 8 | 14 | 0 | 4 | 22 | 48 |
| Total | 1,114 | 837 | 1,716 | 760 | 3,253 | 7,680 |
Points to read from the table:
- Four states hold most of the vacancies. Maharashtra, Tamil Nadu, Odisha and Karnataka together have 4,988 of the 7,680 regular vacancies, about 65% (arithmetic).
- Several large states are missing from this advertisement. There are no rows for Andhra Pradesh, Bihar, Delhi, Haryana, Jharkhand, Kerala, Punjab or Uttarakhand, among others. Exam centres exist in many of those states, but you can only apply for a state that has vacancies.
- Small numbers mean a volatile line. Himachal Pradesh has 9 regular vacancies and Ladakh 7. With so few seats, the last selected mark can swing widely from cycle to cycle.
- PwBD, ex-servicemen and disabled ex-servicemen vacancies are horizontal: they are counted inside the category figures above, not added to them.
Backlog vacancies in the same advertisement: 287 for SC/ST/OBC (SC 58, ST 161, OBC 68), 355 for PwBD and 1,444 for the ex-servicemen categories (964 XS and 480 DXS). Added together that is 2,086, and with the regular vacancies 9,766. Both sums are our arithmetic; the advertisement prints the components. It also notes that these backlog figures are net of the vacancies advertised separately in the special drive described below.
The previous cycle and the special drive
| Recruitment | Advertisement | Regular vacancies | What has happened |
|---|---|---|---|
| Junior Associates 2026-27 | CRPD/CR/2026-27/17, 11.08.2026 | 7,680 | Prelims tentatively in September 2026, Mains tentatively in November 2026 per the advertisement; careers page showed the Prelims call letter stage on 13.11.2026 |
| Special drive for SC/ST/OBC backlog 2026-27 | CRPD/CR/SPLDRIVE/2026-27/16, 07.08.2026 | 1,538 backlog (SC 207, ST 1,219, OBC 112) | Prelims held in September 2026 and individual marks released; careers page notice gives the Mains date as 29.10.2026 |
| Junior Associates 2025-26 | CRPD/CR/2025-26/06, 06.08.2025 | 5,180 (SC 788, ST 450, OBC 1,179, EWS 508, General 2,255) | Prelims held in September 2025, Mains on 21 November 2025, final result and first wait list published |
Regular vacancies rose from 5,180 to 7,680, about 48% more (arithmetic). More vacancies in a state, with other things equal, lower the mark at which the last candidate is shortlisted. Treat that as a reasoning aid and not a forecast: the number of applicants in your state, the difficulty of the paper and the spread of vacancies across categories all move the line too, and the distribution by state changed a great deal between the two years.
How to use this page
These are our suggestions, not SBI's.
- Look up your own row. Your competition is the candidates who applied to your state in your category. The national total tells you little.
- Estimate your shortlist size, not a mark. Multiply your category's vacancies in your state by ten. That is roughly how many will sit Mains from your list.
- Do not chase a printed "safe score". No official one exists. Anyone quoting one is estimating.
- Use your percentile in mocks. On Pareeksha everyone in a cohort sits the same fixed paper, so your percentile is a like-for-like comparison with other candidates. See how fixed-paper percentile works. A percentile tells you more about a rank-based shortlist than a raw score does.
- Remember that Prelims is not the goal. Only Mains marks count. See the exam pattern and preparation strategy.
- If you sat the exam, keep your own score card. It is the only official place your marks appear.
What this page does not have
- Any cut-off marks, for Prelims or Mains, for any state, category or year. We found no SBI document that states them, so none is reproduced or estimated.
- The minimum qualifying percentage on the Mains aggregate. SBI says the Bank decides it and does not print it.
- The Prelims result of the 2026 cycle. It was not on the careers page on 13 November 2026.
- The number of candidates who applied, appeared or were shortlisted, in total or by state. We did not find these figures in an SBI document.
- The state-wise split of backlog, PwBD and ex-servicemen vacancies, and the full state-wise table of the 2025 cycle. They are in the advertisements linked in the sources; we give totals.
Related: Syllabus · Exam pattern · Preparation strategy