IBM model product
2 3B is a compact, text-only decoder-only reasoning LLM with native think, non-think, and low-effort modes, tool calling, and a 128K context window extendable to 512K.
Updated Oct 23, 2026. Default version: IBM Granite 4.2 3B
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for IBM Granite 4.2 3B.
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkProfBench | Score32.10% | Rank04 | Participants4 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkRULER 128k | Score55.30% | Rank04 | Participants4 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkRULER 64k | Score67.52% | Rank04 | Participants4 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkTAU3-Bench | Score45.78% | Rank07 | Participants8 | Percentile14.29% | EvidenceC | Evaluated |
| BenchmarkBFCL-V4 | Score52.41% | Rank17 | Participants22 | Percentile23.81% | EvidenceC | Evaluated |
| BenchmarkArena-Hard v2 | Score34.96% | Rank19 | Participants19 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkHMMT25 | Score66.67% | Rank20 | Participants28 | Percentile29.63% | EvidenceC | Evaluated |
| BenchmarkIFBench | Score74.33% | Rank23 | Participants46 | Percentile51.11% | EvidenceC | Evaluated |
| BenchmarkSciCode | Score24.11% | Rank31 | Participants31 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench v6 | Score69.71% | Rank42 | Participants67 | Percentile37.88% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score78.33% | Rank78 | Participants122 | Percentile36.36% | EvidenceC | Evaluated |
| BenchmarkAA SciCode Subtasks | Score25.35% | Rank82 | Participants89 | Percentile7.95% | EvidenceB | EvaluatedN/A |
| BenchmarkMMLU-Pro | Score67.84% | Rank102 | Participants142 | Percentile28.37% | EvidenceC | Evaluated |
| BenchmarkAA CritPt | Score0.00% | Rank129 | Participants200 | Percentile35.68% | EvidenceB | EvaluatedN/A |
| BenchmarkLM Arena Text Factuality | Score1,324.65 rating | Rank131 | Participants131 | Percentile0.00% | EvidenceA | Evaluated |
| BenchmarkAA HLE Text No Tools | Score6.63% | Rank145 | Participants200 | Percentile27.64% | EvidenceB | EvaluatedN/A |
| BenchmarkAA LCR v1.1 | Score24.33% | Rank156 | Participants193 | Percentile19.27% | EvidenceB | EvaluatedN/A |
| BenchmarkAA GPQA Diamond | Score55.86% | Rank167 | Participants199 | Percentile16.16% | EvidenceB | EvaluatedN/A |
| BenchmarkLM Arena Text | Score1,293.82 rating | Rank173 | Participants218 | Percentile20.74% | EvidenceA | Evaluated |
| BenchmarkGPQA | Score54.80% | Rank188 | Participants256 | Percentile26.67% | EvidenceC | Evaluated |
| BenchmarkAA Omniscience Accuracy | Score9.20% | Rank190 | Participants201 | Percentile5.50% | EvidenceB | EvaluatedN/A |
| BenchmarkLM Arena Text Style Control | Score1,290.55 rating | Rank193 | Participants218 | Percentile11.52% | EvidenceA | Evaluated |
Preference and agent-evaluation results for the default version.
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext factuality | Categoryindustry entertainment and sports and media | Rank120 | Rating / score1,279.49 | Votes758 | ObservationsN/A | Result date |
| Arenatext factuality | Categorycoding | Rank122 | Rating / score1,415.58 | Votes831 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryhard prompts english | Rank124 | Rating / score1,384.15 | Votes846 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry writing and literature and language | Rank124 | Rating / score1,275.65 | Votes825 | ObservationsN/A | Result date |
| Arenatext factuality | Categorylonger query | Rank125 | Rating / score1,340.24 | Votes1,355 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry software and it services | Rank126 | Rating / score1,391.59 | Votes1,193 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryinstruction following | Rank126 | Rating / score1,329.32 | Votes1,093 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryenglish | Rank130 | Rating / score1,354.73 | Votes1,367 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryexclude ties | Rank131 | Rating / score1,291.53 | Votes2,566 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryhard prompts | Rank131 | Rating / score1,355.18 | Votes2,015 | ObservationsN/A | Result date |
| Arenatext factuality | Categorynon english | Rank131 | Rating / score1,307.30 | Votes1,886 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryoverall | Rank131 | Rating / score1,324.65 | Votes3,253 | ObservationsN/A | Result date |
| Arenatext | Categorycoding | Rank151 | Rating / score1,362.64 | Votes831 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry software and it services | Rank157 | Rating / score1,351.17 | Votes1,193 | ObservationsN/A | Result date |
| Arenatext | Categoryexpert | Rank160 | Rating / score1,315.22 | Votes314 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts english | Rank161 | Rating / score1,337.50 | Votes846 | ObservationsN/A | Result date |
| Arenatext style control | Categorycoding | Rank163 | Rating / score1,379.74 | Votes831 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry legal and government | Rank164 | Rating / score1,317.74 | Votes255 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry business and management and financial operations | Rank166 | Rating / score1,287.49 | Votes641 | ObservationsN/A | Result date |
| Arenatext | Categoryenglish | Rank167 | Rating / score1,322.09 | Votes1,367 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank167 | Rating / score1,307.03 | Votes2,015 | ObservationsN/A | Result date |
| Arenatext style control | Categoryexpert | Rank167 | Rating / score1,330.80 | Votes314 | ObservationsN/A | Result date |
| Arenatext | Categoryexclude ties | Rank170 | Rating / score1,240.99 | Votes2,566 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry life and physical and social science | Rank170 | Rating / score1,303.56 | Votes499 | ObservationsN/A | Result date |
| Arenatext | Categoryinstruction following | Rank171 | Rating / score1,274.10 | Votes1,093 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry software and it services | Rank171 | Rating / score1,356.93 | Votes1,193 | ObservationsN/A | Result date |
| Arenatext | Categorylonger query | Rank172 | Rating / score1,292.48 | Votes1,355 | ObservationsN/A | Result date |
| Arenatext | Categorymulti turn | Rank173 | Rating / score1,291.84 | Votes529 | ObservationsN/A | Result date |
| Arenatext | Categoryoverall | Rank173 | Rating / score1,293.82 | Votes3,253 | ObservationsN/A | Result date |
| Arenatext | Categorychinese | Rank174 | Rating / score1,270.37 | Votes206 | ObservationsN/A | Result date |
Official vendor API pricing appears first, followed by individual provider offers.
The default version has no current input or output token prices.
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Provider-specific output speed and catalog latency for IBM Granite 4.2 3B. Runtime does not affect the capability score.
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Technical details for the model's default version.
Available versions of this model. The score column identifies the version used in the overall ranking.
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionIBM Granite 4.2 3B | Released | LLMBoard23.31 | Parameters3B | Context131.1K | Max output131.1K | Open weightsYes | LicenseApache 2.0 |
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Key information about IBM Granite 4.2 3B and its available data.
2 family. 0 and is text-only, with native think, non-think, and low-effort modes, tool calling, and a 128K native context window extendable to 512K.
Data as of 2026-10-22.
Common questions about IBM Granite 4.2 3B.
IBM Granite 4.2 3B's default version was released on Aug 25, 2026.
No official standard PAYG price is currently available for IBM Granite 4.2 3B.
IBM Granite 4.2 3B was created by IBM.
The default version has a 131.1K token context window.
Yes. The default version is marked as open weight under Apache 2.0.
No provider offering is currently linked to the default version.