Updated 1 h ago · first seen 11 Sept 2026
model_01M294YP64QMTD2CBR3K9Y61G0
- Parameters
- 2.31B
- T2 · 2 h ago
- Released
- 16 Apr 2026
- T2 · 1 h ago
- License
- apache-2.0
- T2 · 1 h ago
Specification
- Release date
- 16 Apr 2026
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Openness
- open-weights
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- License
- apache-2.0
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Architecture
- GraniteSpeechForConditionalGeneration
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Parameters
- 2.31B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- Languages
- multilingual, en, fr, de, es, pt, ja
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Base model
- ibm-granite/granite-4.0-1b-base
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- File size
- 4.6 GB
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Hugging Face repo
- ibm-granite/granite-speech-4.1-2b
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- Pipeline tag
- automatic-speech-recognition
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- Model card
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- Downloads
- 359,697
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Likes
- 158
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Gated
- No
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Last modified
- 2026-06-12T17:45:27+00:00
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- Library name
- transformers
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Downloads all time
- 1,964,942
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Model type
- granite_speech
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Tags
- transformers, safetensors, granite_speech, automatic-speech-recognition, multilingual, en, fr, de, es, pt, ja
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Weights dtype
- BF16
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →
Provenance
Attributed facts
20
Source tiers
T220
Freshest observation
1 h ago
Conflicts
None
Modalities
Modalities unavailable.
Capabilities
Tool calling
Unavailable
Structured output
Unavailable
Reasoning
Unavailable
Vision
Unavailable
Audio
Unavailable
Fine-tuning available
Unavailable
No capability flags have been observed from a source yet — we do not infer them.
- Languages
- multilingual, en, fr, de, es, pt, ja
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
No benchmark results recorded
Current prices
No current prices recorded
Price history
Memory need = bytes per parameter (4-bit ≈ 0.5 × 1.15 overhead, 8-bit 1.0, fp16 2.0) + a KV-cache allowance. Not a measurement.
| Hardware | Quantization | Memory | Est. need | Fits |
|---|---|---|---|---|
| NVIDIA DGX B200 | 4bit | 1,440 GB | 1.8 GB | Yes |
| Apple M3 Ultra | 4bit | — | 1.8 GB | Yes |
| Mac Studio (Apple M5 Ultra) | 4bit | — | 1.8 GB | Yes |
| AMD Instinct MI325X | 4bit | 256 GB | 1.8 GB | Yes |
| AMD Instinct MI300X | 4bit | 192 GB | 1.8 GB | Yes |
| Apple M2 Ultra | 4bit | — | 1.8 GB | Yes |
| NVIDIA B200 | 4bit | 180 GB | 1.8 GB | Yes |
| NVIDIA H200 | 4bit | 141 GB | 1.8 GB | Yes |
| NVIDIA H200 NVL | 4bit | 141 GB | 1.8 GB | Yes |
| Apple M1 Ultra | 4bit | — | 1.8 GB | Yes |
| Apple M3 Max | 4bit | — | 1.8 GB | Yes |
| Apple M4 Max | 4bit | — | 1.8 GB | Yes |
| Mac Studio (Apple M5 Max) | 4bit | — | 1.8 GB | Yes |
| MacBook Pro (Apple M5 Max) | 4bit | — | 1.8 GB | Yes |
| NVIDIA DGX Spark | 4bit | 128 GB | 1.8 GB | Yes |
| Apple M2 Max | 4bit | — | 1.8 GB | Yes |
| NVIDIA H100 NVL | 4bit | 94 GB | 1.8 GB | Yes |
| NVIDIA A100 80GB | 4bit | 80 GB | 1.8 GB | Yes |
| NVIDIA H100 SXM | 4bit | 80 GB | 1.8 GB | Yes |
| Apple M1 Max | 4bit | — | 1.8 GB | Yes |
| Apple M4 Pro | 4bit | — | 1.8 GB | Yes |
| Mac mini (Apple M5 Pro) | 4bit | — | 1.8 GB | Yes |
| MacBook Pro (Apple M5 Pro) | 4bit | — | 1.8 GB | Yes |
| Apple M3 Pro | 4bit | — | 1.8 GB | Yes |
| Apple M1 Pro | 4bit | — | 1.8 GB | Yes |
| Apple M2 Pro | 4bit | — | 1.8 GB | Yes |
| Apple M4 | 4bit | — | 1.8 GB | Yes |
| iMac (Apple M4) | 4bit | — | 1.8 GB | Yes |
| Mac mini (Apple M6) | 4bit | — | 1.8 GB | Yes |
| MacBook Air (Apple M5) | 4bit | — | 1.8 GB | Yes |
| MacBook Pro (Apple M5) | 4bit | — | 1.8 GB | Yes |
| NVIDIA GeForce RTX 5090 | 4bit | 32 GB | 1.8 GB | Yes |
| Apple M2 | 4bit | — | 1.8 GB | Yes |
| Apple M3 | 4bit | — | 1.8 GB | Yes |
| NVIDIA GeForce RTX 3090 | 4bit | 24 GB | 1.8 GB | Yes |
| NVIDIA GeForce RTX 4090 | 4bit | 24 GB | 1.8 GB | Yes |
| Apple M1 | 4bit | — | 1.8 GB | Yes |
Ancestors 1
This model
granite-speech-4.1-2b
2.31B params
Quantizations 0
None recorded.
Descendants 0
None recorded.
Papers 7
- arXiv:2310.02943Active35
- arXiv:2505.08699Active35
- arXiv:2603.11243Active35
- arXiv:2604.12398Active35
- arXiv:2604.22817Active35
- arXiv:2604.11269Active35
- arXiv:2603.08397Active35
Repositories 0
No repositories linked yet.
As of
Rewind the record: see this entity's attributes exactly as AI Atlas knew them on a given day.
Claim history · Weights dtype
Weights dtypeweights_dtype1
| Value | Valid from → to | Status | Source | Confidence | Extractor |
|---|---|---|---|---|---|
| BF16 | → current | current | Hugging Face Hub (public pages, model cards, papers)T2 | medium | deterministic |
Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →
| Source | Document | Type | Tier | Last observed | Snapshots |
|---|---|---|---|---|---|
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/ibm-granite/granite-speech-4.1-2b/raw/main/README.md | model_card | T2· Quality secondary | 30 min ago | 1 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/ibm-granite/granite-speech-4.1-2b | model_page | T2· Quality secondary | 1 h ago | 2 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/models?author=ibm-granite&p=0&sort=downloads | listing | T2· Quality secondary | 2 h ago | 2 |
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.