Updated 16 min ago · first seen 11 Sept 2026
model_01M2950GWC93K7X1ZMS20FQCTF
- Parameters
- 25.2B · 4B active
- T2 · 16 min ago
- Released
- 1 Apr 2026
- T2 · 59 min ago
- License
- apache-2.0
- T2 · 59 min ago
Specification
- Release date
- 1 Apr 2026
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min 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 59 min agomedium
- Architecture
- Gemma4ForConditionalGeneration
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Parameters
- 25.2B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium
- Active parameters
- 4B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Base model
- google/gemma-4-26B-A4B-it
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Quantization format
- gguf
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Hugging Face repo
- unsloth/gemma-4-26B-A4B-it-GGUF
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- Pipeline tag
- image-text-to-text
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
- 671,676
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium
- Likes
- 1,115
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 16 min agomedium
- Gated
- No
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Is quantized
- Yes
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Last modified
- 2026-07-17T13:00:49+00:00
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 2 h agomedium
- License url
- https://ai.google.dev/gemma/docs/gemma_4_license
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Downloads all time
- 10,729,859
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Model type
- gemma4
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min agomedium
- Tags
- gguf, gemma4, unsloth, gemma, google, image-text-to-text, imatrix
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 59 min 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
16 min 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.
No structured attributes yet.
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 | 15.0 GB | Yes |
| Apple M3 Ultra | 4bit | — | 15.0 GB | Yes |
| Mac Studio (Apple M5 Ultra) | 4bit | — | 15.0 GB | Yes |
| AMD Instinct MI325X | 4bit | 256 GB | 15.0 GB | Yes |
| AMD Instinct MI300X | 4bit | 192 GB | 15.0 GB | Yes |
| Apple M2 Ultra | 4bit | — | 15.0 GB | Yes |
| NVIDIA B200 | 4bit | 180 GB | 15.0 GB | Yes |
| NVIDIA H200 | 4bit | 141 GB | 15.0 GB | Yes |
| NVIDIA H200 NVL | 4bit | 141 GB | 15.0 GB | Yes |
| Apple M1 Ultra | 4bit | — | 15.0 GB | Yes |
| Apple M3 Max | 4bit | — | 15.0 GB | Yes |
| Apple M4 Max | 4bit | — | 15.0 GB | Yes |
| Mac Studio (Apple M5 Max) | 4bit | — | 15.0 GB | Yes |
| MacBook Pro (Apple M5 Max) | 4bit | — | 15.0 GB | Yes |
| NVIDIA DGX Spark | 4bit | 128 GB | 15.0 GB | Yes |
| Apple M2 Max | 4bit | — | 15.0 GB | Yes |
| NVIDIA H100 NVL | 4bit | 94 GB | 15.0 GB | Yes |
| NVIDIA A100 80GB | 4bit | 80 GB | 15.0 GB | Yes |
| NVIDIA H100 SXM | 4bit | 80 GB | 15.0 GB | Yes |
| Apple M1 Max | 4bit | — | 15.0 GB | Yes |
| Apple M4 Pro | 4bit | — | 15.0 GB | Yes |
| Mac mini (Apple M5 Pro) | 4bit | — | 15.0 GB | Yes |
| MacBook Pro (Apple M5 Pro) | 4bit | — | 15.0 GB | Yes |
| Apple M3 Pro | 4bit | — | 15.0 GB | Yes |
| Apple M1 Pro | 4bit | — | 15.0 GB | Yes |
| Apple M2 Pro | 4bit | — | 15.0 GB | Yes |
| Apple M4 | 4bit | — | 15.0 GB | Yes |
| iMac (Apple M4) | 4bit | — | 15.0 GB | Yes |
| Mac mini (Apple M6) | 4bit | — | 15.0 GB | Yes |
| MacBook Air (Apple M5) | 4bit | — | 15.0 GB | Yes |
| MacBook Pro (Apple M5) | 4bit | — | 15.0 GB | Yes |
| NVIDIA GeForce RTX 5090 | 4bit | 32 GB | 15.0 GB | Yes |
| Apple M2 | 4bit | — | 15.0 GB | Yes |
| Apple M3 | 4bit | — | 15.0 GB | Yes |
| NVIDIA GeForce RTX 3090 | 4bit | 24 GB | 15.0 GB | Yes |
| NVIDIA GeForce RTX 4090 | 4bit | 24 GB | 15.0 GB | Yes |
| Apple M1 | 4bit | — | 15.0 GB | No |
Ancestors 2
- Gemma 4 26B A4B25.8B
- gemma-4-26B-A4BGoogle
This model
unsloth/gemma-4-26B-A4B-it-GGUF
25.2B params
Quantizations 0
None recorded.
Descendants 0
None recorded.
Papers 0
No papers linked yet.
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 · Parameters
Parametersparameter_count3
| Value | Valid from → to | Status | Source | Confidence | Extractor |
|---|---|---|---|---|---|
| 25.2B | → current | current | Hugging Face Hub (public pages, model cards, papers)T2 | medium | deterministic |
| 26B | → | superseded | Hugging Face Hub (public pages, model cards, papers)T2 | medium | deterministic |
| 25.2B | → | superseded | 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 →
unsloth/gemma-4-26B-A4B-it-GGUF: parameter count changed from 26000000000 to 25233142046
Parameters26B→25.2Bhuggingfaceunsloth/gemma-4-26B-A4B-it-GGUF: parameter count changed from 25233142046 to 26000000000
Parameters25.2B→26BhuggingfaceNew model: unsloth/gemma-4-26B-A4B-it-GGUF (Unsloth)
huggingface
| Source | Document | Type | Tier | Last observed | Snapshots |
|---|---|---|---|---|---|
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/models?author=unsloth&p=0&sort=downloads | listing | T2· Quality secondary | 16 min ago | 3 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/unsloth/gemma-4-26B-A4B-it-GGUF/raw/main/README.md | model_card | T2· Quality secondary | 34 min ago | 1 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/unsloth/gemma-4-26B-A4B-it-GGUF | model_page | T2· Quality secondary | 59 min ago | 1 |
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.