Step-3.5-Flash-REAP-121B-A11B
Cerebras Systemshuggingface.co/cerebras/Step-3.5-Flash-REAP-121B
Updated 1 h ago · first seen 11 Sept 2026
model_01M294XYQYY7Q5X9ZA902FHNWP
- Parameters
- 121B · 11B active
- T2 · 1 h ago
- Released
- 25 Feb 2026
- T2 · 1 h ago
- License
- other
- T2 · 1 h ago
Specification
- Release date
- 25 Feb 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 1 h agomedium
- License
- other
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Architecture
- Step3p5ForCausalLM
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Parameters
- 121B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Active parameters
- 11B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Languages
- en
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Base model
- stepfun-ai/Step-3.5-Flash
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- File size
- 241.9 GB
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Hugging Face repo
- cerebras/Step-3.5-Flash-REAP-121B-A11B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Pipeline tag
- text-generation
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Model card
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Downloads
- 148
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Likes
- 15
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
- 2026-02-25T07:13:23+00:00
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Library name
- transformers
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Downloads all time
- 943
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Model type
- step3p5
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Tags
- transformers, safetensors, step3p5, text-generation, stepfun, MOE, pruning, compression, custom_code, en
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 1 h agomedium
- Weights dtype
- BF16, F32
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
21
Source tiers
T221
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
- en
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 |
|---|---|---|---|---|
| Apple M3 Ultra | 4bit | — | 70.1 GB | Yes |
| AMD Instinct MI325X | 4bit | 256 GB | 70.1 GB | Yes |
| AMD Instinct MI300X | 4bit | 192 GB | 70.1 GB | Yes |
| Apple M2 Ultra | 4bit | — | 70.1 GB | Yes |
| NVIDIA B200 | 4bit | 180 GB | 70.1 GB | Yes |
| NVIDIA H200 | 4bit | 141 GB | 70.1 GB | Yes |
| Apple M1 Ultra | 4bit | — | 70.1 GB | Yes |
| Apple M3 Max | 4bit | — | 70.1 GB | Yes |
| Apple M4 Max | 4bit | — | 70.1 GB | Yes |
| NVIDIA DGX Spark | 4bit | 128 GB | 70.1 GB | Yes |
| Apple M2 Max | 4bit | — | 70.1 GB | Yes |
| NVIDIA A100 80GB | 4bit | 80 GB | 70.1 GB | Yes |
| NVIDIA H100 SXM | 4bit | 80 GB | 70.1 GB | Yes |
| Apple M1 Max | 4bit | — | 70.1 GB | No |
| Apple M4 Pro | 4bit | — | 70.1 GB | No |
| Apple M3 Pro | 4bit | — | 70.1 GB | No |
| Apple M1 Pro | 4bit | — | 70.1 GB | No |
| Apple M2 Pro | 4bit | — | 70.1 GB | No |
| Apple M4 | 4bit | — | 70.1 GB | No |
| NVIDIA GeForce RTX 5090 | 4bit | 32 GB | 70.1 GB | No |
| Apple M2 | 4bit | — | 70.1 GB | No |
| Apple M3 | 4bit | — | 70.1 GB | No |
| NVIDIA GeForce RTX 3090 | 4bit | 24 GB | 70.1 GB | No |
| NVIDIA GeForce RTX 4090 | 4bit | 24 GB | 70.1 GB | No |
| Apple M1 | 4bit | — | 70.1 GB | No |
Ancestors 1
- Step 3.5 FlashStepFun
This model
Step-3.5-Flash-REAP-121B-A11B
121B params
Quantizations 0
None recorded.
Descendants 0
None recorded.
Papers 1
- arXiv:2510.13999Active35
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 · Base model
Base modelbase_model1
| Value | Valid from → to | Status | Source | Confidence | Extractor |
|---|---|---|---|---|---|
| stepfun-ai/Step-3.5-Flash | → 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 →
New model: Step-3.5-Flash-REAP-121B-A11B (Cerebras Systems)
huggingface
| Source | Document | Type | Tier | Last observed | Snapshots |
|---|---|---|---|---|---|
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/cerebras/Step-3.5-Flash-REAP-121B-A11B/raw/main/README.md | model_card | T2· Quality secondary | 18 min ago | 1 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/cerebras/Step-3.5-Flash-REAP-121B-A11B | model_page | T2· Quality secondary | 1 h ago | 2 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/models?author=cerebras&p=0&sort=downloads | listing | T2· Quality secondary | 1 h ago | 2 |
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.