Project AURIONAI Use-Cases Research & Innovation
IBM

IBM

Big tech, USA. 57 models in the directory.

big techUSA
0deployments

Identity

Type
big tech
Country label
USA
HQ country
USA
Founding country
USA
Parent
not in the notes
Parent country
not in the notes
Country basis
the research notes
Where this comes from
Convention D1: a dual label, HQ first, when the HQ, founding or parent country differ.

Models by Class

LLM13

Models

ModelClassOpennessParametersReleased
IBMGranite 4.2 30BLLMopen source30BSep 2026IBMGranite 4.2 3BLLMopen source3BSep 2026IBMGranite 4.2 8BLLMopen source8BSep 2026IBMGranite TimeSeries TTM r3time seriesopen weights permissive–21 May 2026IBMGranite Embedding Multilingual R2embeddingopen weights permissive–14 May 2026IBMgranite-speech-4.1-2bspeech recognitionopen weights permissive2.31B16 Apr 2026IBMGranite Guardian 4.1-8Bguardopen weights permissive8BApr 2026IBMGranite PatchTST-FM r1IBM, Rensselaer Polytechnic Institutetime seriesopen weights permissive257.9M11 Mar 2026IBMgranite-4.0-1b-speechspeech recognitionopen weights permissive2.31B27 Feb 2026IBMGranite 4.0LLMopen weights permissive–2 Oct 2025IBMGranite-4.0-H-Microedge LLMopen weights permissive3B2 Oct 2025IBMFlowState r1.1time seriesopen weights permissive9.07M10 Sep 2025IBMSuryaIBM, NASAscientificopen weights permissive–13 Aug 2025IBMGranite Embedding Reranker English R2rerankeropen weights permissive149.61M4 Aug 2025IBMGranite Embedding small English R2embeddingopen weights permissive47.66M17 Jul 2025IBMGranite Vision 3.3 2BOCR and documentopen weights permissive2.98B11 Jun 2025IBMGranite TSPulsetime seriesopen weights permissive1.08M19 May 2025IBMGranite 3.3 8B InstructLLMopen weights permissive8.17B9 Apr 2025IBMGranite 3.2 8B InstructLLMopen weights permissive8.17B17 Feb 2025IBMTerraMindIBM, European Space Agency (ESA), Julich (Forschungszentrum Julich)geospatialopen source–14 Feb 2025IBMgranite-speech-3.3-8bspeech recognitionopen weights permissive8B2025IBMTerraCodecgeospatialopen source–2025IBMGranite 3.1 8B InstructLLMopen weights permissive8.17B6 Dec 2024IBMGranite Embeddingembeddingopen weights permissive–4 Dec 2024IBMGranite Embedding 278M multilingualembeddingopen weights permissive278.04M4 Dec 2024IBMPrithvi-EO-2.0IBM, NASA, Julich (Forschungszentrum Julich)geospatialopen source–3 Dec 2024IBMPrithvi-EO-2.0 100M-TLIBM, NASAgeospatialopen source100M3 Dec 2024IBMPrithvi-EO-2.0 300MIBM, NASAgeospatialopen source300M3 Dec 2024IBMPrithvi-EO-2.0 600MIBM, NASAgeospatialopen source600M3 Dec 2024IBMPrithvi-EO-2.0-600M-TLIBM, NASAgeospatialopen weights permissive600M3 Dec 2024IBMPrithvi-EO-2.0 TINY-TLIBM, NASAgeospatialopen source5M3 Dec 2024IBMGranite 3LLMopen weights permissive–2 Oct 2024IBMGranite 3.0 8B InstructLLMopen weights permissive8.17B2 Oct 2024IBMPrithvi-WxCIBM, NASAweatheropen weights permissive2.3B20 Sep 2024IBMGranite 8B Code Instruct 128Kcodeopen weights permissive8.05B12 Jul 2024IBMGranite Codecodeopen weights permissive–26 Apr 2024IBMGranite 20B Code Instruct 8Kcodeopen weights permissive20.07B26 Apr 2024IBMGranite 34B Code Instruct 8Kcodeopen weights permissive33.72B26 Apr 2024IBMGranite TinyTimeMixer (TTM) r2time seriesopen weights permissive805K2024IBMPrithvi-EO 1.0IBM, NASAgeospatialunknown–2023Princeton UniversityPatchTSTPrinceton University, IBMtime seriesopen source–27 Nov 2022IBMCrossViTvisionunknown–27 Mar 2021IBMGranite-4.0-H-Micro-BaseLLMunknown––IBMgranite-4.1-8bLLMunknown8B–IBMGranite-4.1-8B-BaseLLMunknown8B–IBMGranite 4.2LLMopen source––IBMGranite 4 350Medge LLMunknown––IBMGranite FlowStatetime seriesunknown––IBMGranite-Geospatial-Oceangeospatialunknown––IBMGranite Guardian 3.3-8Bguardopen weights permissive8B–IBMGranite Guardian 4.0-3B toxicity-jaguardopen weights permissive3B–IBMGranite Guardian HAP 125Mguardopen weights permissive125M–IBMGranite Guardian HAP 38Mguardopen weights permissive38M–IBMGranite PatchTST-FM r2IBM, Rensselaer Polytechnic Institutetime seriesunknown––IBMGranite TTM-R1time seriesunknown––IBMGranite TinyTimeMixer (TTM) r2.1time seriesopen weights permissive––IBMPrithvi v1IBM, NASAgeospatialunknown––

Scores

BenchmarkModelScoreWho ran itSetting
BFCLGranite-4.0-H-Micro57.56 %vendor-reportedBFCL v3
FC Reward BenchGranite Guardian 4.1-8B0.79 F1vendor-reported
GIFT-EvalFlowState r1.10.487 CRPS (geo-mean, normalised to Seasonal Naive)independentzero-shot; rank is average CRPS rank
GIFT-EvalFlowState r1.10.701 MASE (geo-mean, normalised to Seasonal Naive)independentzero-shot
GIFT-EvalFlowState r1.125.6 mean CRPS rankvendor-reportedzero-shot foundation-model CRPS rank, lower is better
GIFT-EvalGranite FlowStaterank onlyvendor-reportedIBM Granite docs
GIFT-EvalGranite PatchTST-FM r123.1 mean CRPS rankvendor-reportedzero-shot foundation-model CRPS rank, lower is better
GIFT-EvalGranite PatchTST-FM r20.467 CRPS (geo-mean, normalised to Seasonal Naive)independentzero-shot; rank is average CRPS rank
GIFT-EvalGranite PatchTST-FM r20.685 MASE (geo-mean, normalised to Seasonal Naive)independentzero-shot
GIFT-EvalGranite TinyTimeMixer (TTM) r20.873 CRPS (geo-mean, normalised to Seasonal Naive)independentpretrained, test leakage Yes; rank is average CRPS rank
GIFT-EvalGranite TinyTimeMixer (TTM) r21.02 MASE (geo-mean, normalised to Seasonal Naive)independentpretrained, test leakage Yes
GPQAGranite-4.0-H-Micro32.15 %vendor-reported0-shot, CoT
Granite Guardian out-of-distribution safety setGranite Guardian 4.1-8B0.79 F1vendor-reportedout-of-distribution safety
GSM8KGranite-4.0-H-Micro81.35 %vendor-reported8-shot
Hugging Face Open ASR Leaderboardgranite-4.0-1b-speech5.52 WERmaintainer-runeight-set macro-average English WER
Hugging Face Open ASR Leaderboardgranite-speech-3.3-8b5.74 WERmaintainer-runeight-set macro-average English WER
Hugging Face Open ASR Leaderboardgranite-speech-4.1-2b5.33 WERmaintainer-runeight-set macro-average English WER
HumanEvalGranite-4.0-H-Micro81 %vendor-reportedpass@1
IFBenchGranite 4.2 8B79.33 %vendor-reported
IFEvalGranite-4.0-H-Micro84.32 %vendor-reportedAverage of instruct and prompt strict
LM-AggreFactGranite Guardian 4.1-8B0.76 F1vendor-reported
MMLUGranite-4.0-H-Micro67.43 %vendor-reported5-shot
MMLU-ProGranite 4.2 8B74.04 %vendor-reported
MMLU-ProGranite-4.0-H-Micro43.48 %vendor-reported5-shot, CoT
SWE-bench VerifiedGranite 4.2 8B47.67 %vendor-reported
TSB-ADGranite TSPulserank onlyvendor-reportedIBM claims #1 on univariate and multivariate, +20%; no scores published on card; conflicts with PaAno rank 1
Where this comes from
Every recorded score of this organisation's models, with who ran it. Scores from different settings are not directly comparable.

Sources

  1. ibm-granite (IBM Granite) huggingface.co https://huggingface.co/ibm-granite
  2. Open ASR Leaderboard Benchmark Scores & AI Model Leaderboard | BenchmarkList benchmarklist.com https://benchmarklist.com/benchmarks/open_asr_leaderboard/
  3. ibm-granite/granite-timeseries-ttm-r2 , Hugging Face huggingface.co https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2
  4. library ollama.com https://ollama.com/library
  5. Granite Time Series | IBM Granite ibm.com https://www.ibm.com/ae-ar/granite/docs/models/time-series/
  6. ibm-granite/granite-timeseries-tspulse-r1 , Hugging Face huggingface.co https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "IBM", https://aurion.semai.id/organisations/ibm/, records of October 2026.