vs
2:5
overall capability wins (2 ties across 9 scored). Granite 4.2 3B leads overall.
- Reasoning1:3
- Coding0:2
- Specs1:0
Verdict
Granite 4.2 3B leads coding (widest gap: +10.5 on TermBench).
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Granite 4.1 8B
- $0.08
- Granite 4.2 3B
- -
Price per 1M tokens
Full comparison
Organization
IBM
IBM
Family
Granite
Granite
License
Apache 2.0
Apache 2.0
Open weights
Yes
Yes
Release
Apr 30, 2026
Aug 25, 2026
Knowledge cutoff
-
-
API / provider
IBM watsonx
No primary API price listed
Modalities
text → text
text → text
Specs
Context window
131K
131K
Max output
8K
8K
Parameters
8B
3B
Pricing
Input $/1M
$0.05
—
Output $/1M
$0.10
—
Blended $/1M (3∶1)
$0.06
-
Speed
tok/s
117
-
TTFT (s)
0.13
-
Reasoning
MMLU-Pro
—
—
GPQA Diamond
43.3
55.9
Humanity's Last Exam
3.8
6.6
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
12.3
9.2
AA-LCR v1.1
13.3
24.3
CritPt
—
—
MMMU-Pro
—
—
IFBench
38.6
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
3.4
13.9
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
21.8
25.3
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
1,192
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,306
—
| Metric | Granite 4.1 8B | Granite 4.2 3B | Delta |
|---|---|---|---|
| Identity | |||
| Organization | IBM | IBM | — |
| Family | Granite | Granite | — |
| License | Apache 2.0 | Apache 2.0 | — |
| Open weights | Yes | Yes | — |
| Release | Apr 30, 2026 | Aug 25, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | IBM watsonx | No primary API price listed | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 131K | 131K | tie |
| Max output | 8K | 8K | tie |
| Parameters | 8B | 3B | A +5.00 |
| Pricing | |||
| Input $/1M | $0.05 | — | — |
| Output $/1M | $0.10 | — | — |
| Blended $/1M (3∶1) | $0.06 | - | — |
| Speed | |||
| tok/s | 117 | - | — |
| TTFT (s) | 0.13 | - | — |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 43.3 | 55.9 | B +12.6 pts |
| Humanity's Last Exam | 3.8 | 6.6 | B +2.8 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 12.3 | 9.2 | A +3.1 pts |
| AA-LCR v1.1 | 13.3 | 24.3 | B +11.0 pts |
| CritPt | — | — | — |
| MMMU-Pro | — | — | — |
| IFBench | 38.6 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 3.4 | 13.9 | B +10.5 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 21.8 | 25.3 | B +3.5 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | 1,192 | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,306 | — | — |
Granite 4.1 8B
- Context
- 131K / 8K out
- Parameters
- 8B
- Price
- $0.05 / $0.10
- Speed
- 117 tok/s · 0.13s TTFT
- Modalities
- text → text
- License
- Apache 2.0
Token-efficient IBM enterprise 8B with tool calling and fill-in-the-middle code support under Apache 2.0.
Granite 4.2 3B
- Context
- 131K / 8K out
- Parameters
- 3B
- Price
- —
- Speed
- —
- Modalities
- text → text
- License
- Apache 2.0
Compact edge-oriented member of IBM's Granite 4.2 family (Apache 2.0, Aug 25, 2026) — a dense 3B reasoning model with the same native 128K context and switchable thinking mode as its 8B and 30B siblings.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.