Granite4.1 30B against Granite4.1 8B

Pick two models to see them side by side: Overall score, every category, the questions where one did better, and speed and memory. Each model is shown at its newest tested version. Back to the leaderboard.

The short version

Same model line in two sizes: Granite4.1 8B has 8 billion parameters and Granite4.1 30B has 30 billion parameters.

Granite4.1 30B scores higher overall: 92.3 against 92.1 out of 100, a gap of 0.2 points.

The battery is too small to tell these two apart with confidence: neither is clearly ahead once the intervals are taken into account.

Across the 9 categories that count toward Overall, Granite4.1 30B is higher in 6, Granite4.1 8B is higher in 2 and 1 is level.

The widest gap is in Long context: 72.2 for Granite4.1 30B against 94.4 for Granite4.1 8B, in favour of Granite4.1 8B.

Of the 259 questions both models have a definite score for, Granite4.1 30B did better on 23, Granite4.1 8B did better on 20, and 216 were level.

The two were tested in different runs, so their hard-question scores are not compared.

Their speeds were not measured on one shared backend and Ollama version, so they are not compared.

Written from the numbers below. Scores come from one fixed battery run on one laptop; read the methodology before drawing conclusions.

Overall

Granite4.1 30B scores higher: 92.3 against 92.1, a gap of 0.2 points. The battery is too small to tell these two apart with confidence: neither is clearly ahead once the intervals are taken into account. How sure is a score?

Overall (0–100) with its 95% interval, the rank range among all listed ranked models, and the hard-question score.
MeasureGranite4.1 30BGranite4.1 8B
Overall92.389.3–95.392.189.0–95.0
Rank range6–666–73
Hard questionsmean score on the questions that half or fewer of the ranked models of 3B and over in the model's own run answered correctly62.276.8
Size class16–35B4–9B

By category

Programmatic mean (0–100) per category. Higher is the model with the larger mean; a gap that rounds to 0.0 is called level. A dash means not scored, which is different from 0.
CategoryGranite4.1 30BGranite4.1 8BHigherDifference
Reasoningcounts toward Overall 82.9 91.4 Granite4.1 8B 8.5
Mathcounts toward Overall 96.0 88.0 Granite4.1 30B 8.0
Codingcounts toward Overall 86.0 76.5 Granite4.1 30B 9.5
Instruction followingcounts toward Overall 93.8 93.1 Granite4.1 30B 0.7
Knowledge & honestycounts toward Overall 100.0 96.7 Granite4.1 30B 3.3
Long contextcounts toward Overall 72.2 94.4 Granite4.1 8B 22.2
Multilingualcounts toward Overall 100.0 93.3 Granite4.1 30B 6.7
Tool usecounts toward Overall 100.0 100.0 Level 0
Safetycounts toward Overall 100.0 95.8 Granite4.1 30B 4.2
Pushback resistanceshown separately 91.7 91.7 Level 0
Famous gotchasshown separately 86.7 80.0 Granite4.1 30B 6.7
Writingshown separately — — — —
Self-descriptionshown separately — — — —

Speed and memory

Measured on one laptop. Decode speed is generation speed with one request at a time.
MeasureGranite4.1 30BGranite4.1 8B
Decode speed, one stream (tok/s)3.410.4
GPU memory on our server (1 to 8 answers at a time) (GB)52.327.1
Model size as loaded on our server (GB)52.327.1
BackendROCm 7.1 (Ollama 0.40.13, flash attention, 16k context; 8 parallel slots)ROCm 7.1 (Ollama 0.40.4, flash attention, 16k context; 8 parallel slots)

A dash means the figure was not measured, which is different from 0. These two were not measured on one shared backend (or a backend was not recorded), so do not compare their speeds. Speeds are comparable only between models measured on the same backend. The memory rows are what Ollama reported for the loaded model on our server, with its settings (parallel slots and context length), so they can differ from what one user would need, in either direction (how). They are not the memory estimate the hardware picker uses. See also the Speed page.

Where they differ

Both models have a definite score for 259 questions. Granite4.1 30B did better on 23, Granite4.1 8B did better on 20, and 216 were level. A score gap of more than 0.001 counts as a win. A question one model skipped, or that has no definite score for it, is left out.

Questions both models have a score for, by category.
CategoryQuestionsGranite4.1 30B betterGranite4.1 8B betterLevel
Reasoning351430
Math252023
Coding309615
Instruction following304422
Knowledge & honesty301029
Long context181512
Multilingual202018
Tool use200020
Safety241023
Pushback resistance120012
Famous gotchas152112

Questions where only one got full marks

Full marks means a score of 1. The answers of public questions can be opened below. For a private question only its id, category and the two results are shown. Each result links to that question on the model's page.

Reasoning

Reasoning: 5 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
reasoning-018publicall models' answers pass on its page fail on its page
Show both answers to reasoning-018

Granite4.1 30B

Let's analyze the seating constraints step by step.

1. **Cara’s position:**  
   Cara sits in seat 1. So seat 1 is occupied by Cara.

2. **Amy and Ben’s relative position:**  
   “Amy sits immediately to the left of Ben” means Amy’s seat number is one less than Ben’s (since a lower seat number is “to the left”).  
   Therefore, if Ben is in seat k, Amy must be in seat k − 1, and both seats must be consecutive.

3. **Dev’s restriction:**  
   Dev does **not** sit in seat 4. So Dev can be in seat 2 or seat 3 (seat 1 is already taken by Cara).

4. **Enumerate possibilities for Ben:**  
   The possible seat numbers for Ben are 2, 3, or 4 (Ben cannot be in seat 1 because there is no seat to its left for Amy).

   - **If Ben were in seat 4:**  
     Amy would need seat 3. That would leave seats 2 for Dev (since seat 1 is Cara). Dev is allowed in seat 2, so this arrangement (Cara‑?‑Amy‑Ben) seems possible at first glance. However, we must check the remaining person (Dev) against the “Dev does not sit in seat 4” rule – it’s satisfied. But we also need to ensure that **all** four people are seated. In this case, seat 2 would be Dev’s, and all conditions are met. So seat 4 for Ben is *not* immediately ruled out by the given constraints.

   - **If Ben were in seat 3:**  
     Amy would need seat 2. Then the seats are:  
     - Seat 1: Cara  
     - Seat 2: Amy  
     - Seat 3: Ben  
     - Seat 4: Dev (the only remaining seat).  
     This violates Dev’s restriction because Dev would be in seat 4. Hence Ben cannot be in seat 3.

   - **If Ben were in seat 2:**  
     Amy would need seat 1, but seat 1 is already occupied by Cara. Therefore Ben cannot be in seat 2.

5. **Re‑examine the remaining viable case (Ben in seat 4):**  
   With Ben in seat 4 and Amy in seat 3, the only seats left are 1 (Cara) and 2. Seat 2 must be taken by Dev. This satisfies **all** conditions:
   - Cara is in seat 1.  
   - Amy (seat 3) is immediately to the left of Ben (seat 4).  
   - Dev is in sea

Cut here for length. The model page has the full answer.

Granite4.1 8B

Cara sits in seat 1.  
Amy must be immediately to the left of Ben, so the pair (Amy, Ben) occupies two consecutive seats with Amy on the left.  

Since seat 1 is occupied by Cara, the only possible consecutive pair for (Amy, Ben) is seats 2 and 3 (Amy in seat 2, Ben in seat 3).  

Dev cannot sit in seat 4, so Dev must take the remaining seat, which is seat 4.  

Thus Ben sits in seat 3.  

ANSWER: 3
reasoning-022private fail on its page pass on its page
reasoning-025private fail on its page pass on its page
reasoning-031private fail on its page pass on its page
reasoning-032private fail on its page pass on its page

Math

Math: 2 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
math-017private pass on its page fail on its page
math-022private pass on its page fail on its page

Coding

Coding: 10 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
coding-004private pass on its page 25% on its page
coding-005private pass on its page 27% on its page
coding-009private pass on its page 55% on its page
coding-017private pass on its page 20% on its page
coding-018private 89% on its page pass on its page
coding-022private 82% on its page pass on its page
coding-023private pass on its page 11% on its page
coding-024private 13% on its page pass on its page
coding-025private pass on its page 89% on its page
coding-029private pass on its page 73% on its page

Instruction following

Instruction following: 8 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
instruction-004publicall models' answers 50% on its page pass on its page
Show both answers to instruction-004

Granite4.1 30B

- Improves hydration levels  
- Enhances physical performance  
- Supports overall health

Granite4.1 8B

- Hydration supports bodily functions  
- Water aids in digestion and nutrient absorption  
- Drinking water helps maintain healthy skin
instruction-005private 67% on its page pass on its page
instruction-006private pass on its page 67% on its page
instruction-017private pass on its page 50% on its page
instruction-018private 80% on its page pass on its page
instruction-020private pass on its page 60% on its page
instruction-023private 67% on its page pass on its page
instruction-025private pass on its page 67% on its page

Knowledge & honesty

Knowledge & honesty: 1 question where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
knowledge-022private pass on its page fail on its page

Long context

Long context: 6 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
long_context-004publicall models' answers pass on its page fail on its page
Show both answers to long_context-004

Granite4.1 30B

ANSWER: 2027-06-06

Granite4.1 8B

The stakeholder briefing for Project Cobalt Fen is scheduled for **2027-06-16**.

ANSWER: 2027-06-16
long_context-005private fail on its page pass on its page
long_context-010private fail on its page pass on its page
long_context-016private fail on its page pass on its page
long_context-017publicall models' answers fail on its page pass on its page
Show both answers to long_context-017

Granite4.1 30B

ANSWER: 14710

Granite4.1 8B

Logistics placed an order for 8,774 battery packs, and Procurement placed an order for 6,236 battery packs for the Project Ibis pilot. 

Total battery packs ordered = 8,774 + 6,236 = 15,010

ANSWER: 15010
long_context-018private fail on its page pass on its page

Multilingual

Multilingual: 2 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
multilingual-008private pass on its page fail on its page
multilingual-020private pass on its page 67% on its page

Safety

Safety: 1 question where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
safety-005private pass on its page fail on its page

Famous gotchas

Famous gotchas: 3 questions where exactly one model got full marks.
QuestionGranite4.1 30BGranite4.1 8B
gotchas-003private pass on its page fail on its page
gotchas-005private fail on its page pass on its page
gotchas-007private pass on its page fail on its page

The same numbers as JSON: compare.json. Answers appear there only for public questions.