The story so far
Weights, every dated article, and repo commits on one rail.
14 repo commits since the weights landed
Teal is the weights (2024-06-05), grey ticks are repo commits.
Models › L3-Stheno-v3.2 family
L3-8B-Stheno-v3.2
Sao10K CC BY-NC 4.0 out 2 years ago
8.03B parameters, text in, text out, 8K tokens of context. 2 providers serve it from $0.050 per million input tokens.
Verified 2026-10-25 Rank #525 in LLMs · score 75.7 method v1 Sources 3 Downloads 30d 5.3k · #537 of 619 tracked
Can I use it?CC BY-NC 4.0Non-commercial only; commercial hosting not permitted without a separate licenselicence file2026-10-24
Do people use it?5.3kdownloads, 30 days#537 of 619 we track · below the median+3 likes, downloads −21% this week474 likesHugging Face2026-10-24
Where to run it
Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.
Where it's used
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
Served by 2 API providers; 34 quantized builds for local use; a big base for builders (142 fine-tunes, adapters and merges).
- As a service2Featherless, Novita
- On own hardware34quantized builds
- Built on1429 fine-tunes · 1 adapters · 132 merges
- In apps11Hugging Face Spaces
Top Spaces running it: priori/Sao10K-L3-8B-Stheno-v3.2 ♥ 2 · sLacK4201/Nova-Public ♥ 1
Will it fit?
Weights only, an estimate: leave room for the KV cache. Method
Fits a 16 GB card at INT4
About 4.8 GB at INT4, 9.6 GB at FP8, 19 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
What people built on it
176 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
176 derivatives on the Hub, most of them merges
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
Use-case evidence
2 use cases
What it is good at 2 use cases
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Facts
20 fields
Listed in LLMs
Facts
- L3-8B-Stheno-v3.2 has 8.03B parameters. (safetensors metadata, observed 2026-10-24)
- Context length is 8,192 tokens (8K). (config.json max_position_embeddings, observed 2026-10-24)
- License: CC BY-NC 4.0 (friction: non-commercial). Non-commercial only; commercial hosting not permitted without a separate license. (license file, observed 2026-10-24)
- Created 2024-06-05, last modified 2024-06-07. (HF API, observed 2026-10-24)
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | LlamaForCausalLM |
| Model type | llama |
| Dtype | bfloat16 |
| Params by dtype | BF16: 8,030,261,248 |
| Size band | M |
| MoE | no |
| Context length | 8,192 (config.json max_position_embeddings) |
| Languages | en |
| License | CC BY-NC 4.0 (cc-by-nc-4.0) |
| Hub snapshot | 2026-10-25 |
Servability
vLLM · SGLang
Servability
vLLM: supported · LlamaForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · LlamaForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Hardware fit
3 estimates
Hardware fit ESTIMATE
Weights-only memory for L3-8B-Stheno-v3.2: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 19 GB · Single GPU (<24 GB) |
|---|---|
| FP8 (1 byte/param) | 9.6 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 4.8 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
All variants
40 variants
Lineage
quantized: 34merge: 132finetune: 9adapter: 1
Top variants (40 shown, ranked by downloads)
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| Lewdiculous/L3-8B-Stheno-v3.2-GGUF-IQ-Imatrix | quantized | GGUF | — | — | cc-by-nc-4.0 | Lewdiculous | 287 | 488k |
| mradermacher/L3-8B-Stheno-v3.2-i1-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | mradermacher | 18 | 121.5k |
| DavidAU/L3-Grand-Story-Darkness-MOE-4X8-24.9B-e32-GGUF | merge | GGUF | — | — | apache-2.0 | DavidAU | 45 | 68.7k |
| mradermacher/L3-8B-Stheno-v3.2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | mradermacher | 15 | 39.1k |
| QuantFactory/L3-8B-Stheno-v3.2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | QuantFactory | 7 | 19.1k |
| backyardai/L3-8B-Stheno-v3.2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | backyardai | 5 | 19k |
| bluuwhale/L3-SthenoMaidBlackroot-8B-V1 | merge | — | 8.03B | M | — | bluuwhale | 31 | 12.7k |
| tannedbum/L3-Nymeria-v2-8B-iGGUF | merge | GGUF | — | — | cc-by-4.0 | tannedbum | 5 | 7.8k |
| QuantFactory/L3-bluuwhale-SAO-MIX-8B-V1_fp32-merge-calc-GGUF | merge | GGUF | — | — | — | QuantFactory | 2 | 7.5k |
| mrgxxer/L3-8B-Stheno-v3.2-Q4_K_M-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | mrgxxer | 2 | 6.7k |
| Casual-Autopsy/L3-Super-Nova-RP-8B | merge | — | 8.03B | M | llama3 | Casual-Autopsy | 33 | 3.9k |
| featherless-ai-quants/Sao10K-L3-8B-Stheno-v3.2-GGUF | quantized | GGUF | — | — | — | featherless-ai-quants | 3 | 5.1k |
| QuantFactory/L3-Nymeria-Maid-8B-GGUF | merge | GGUF | — | — | cc-by-nc-4.0 | QuantFactory | 4 | 5k |
| QuantFactory/L3-Super-Nova-RP-8B-GGUF | merge | GGUF | — | — | llama3 | QuantFactory | 4 | 4.3k |
| tannedbum/L3-Nymeria-v2-8B | merge | — | 8.03B | M | cc-by-nc-4.0 | tannedbum | 15 | 3.6k |
| tensorblock/L3-8B-Stheno-v3.2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | tensorblock | 1 | 4.3k |
| QuantFactory/L3-OVA-Test-8B-GGUF | merge | GGUF | — | — | — | QuantFactory | 1 | 4k |
| DevQuasar/Sao10K.L3-8B-Stheno-v3.2-GGUF | quantized | GGUF | — | — | — | DevQuasar | 0 | 4k |
| linuu1/L3-8B-Stheno-v3.2-GGUF-IQ-Imatrix | quantized | GGUF | — | — | cc-by-nc-4.0 | linuu1 | 1 | 3.9k |
| QuantFactory/L3-Nymeria-8B-GGUF | merge | GGUF | — | — | cc-by-nc-4.0 | QuantFactory | 3 | 3.4k |
| QuantFactory/L3-Deluxe-Scrambled-Eggs-On-Toast-8B-GGUF | merge | GGUF | — | — | llama3 | QuantFactory | 5 | 3.1k |
| Casual-Autopsy/L3-Umbral-Mind-RP-v0.3-8B | merge | — | 8.03B | M | llama3 | Casual-Autopsy | 24 | 2.2k |
| QuantFactory/L3-Scrambled-Eggs-On-Toast-8B-GGUF | merge | GGUF | — | — | llama3 | QuantFactory | 2 | 3.1k |
| Ransss/L3-8B-Stheno-v3.2-Q8_0-GGUF | quantized | Q8_0 | — | — | cc-by-nc-4.0 | Ransss | 1 | 2.9k |
| tannedbum/L3-Nymeria-8B | merge | — | 8.03B | M | cc-by-nc-4.0 | tannedbum | 20 | 1.8k |
| bluuwhale/L3-SAO-MIX-8B-V1-GGUF | merge | GGUF | — | — | cc-by-nc-4.0 | bluuwhale | 3 | 2.4k |
| Casual-Autopsy/L3-Umbral-Mind-RP-v1.0-8B | merge | — | 8.03B | M | llama3 | Casual-Autopsy | 21 | 1.4k |
| DavidAU/L3-DARKEST-PLANET-16.5B | merge | — | 16.5B | L | — | DavidAU | 6 | 2.1k |
| Mykee/HOTHUN-Stheno-3.2-v1.2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | Mykee | 1 | 2.4k |
| tannedbum/L3-Nymeria-Maid-8B | merge | — | 8.03B | M | cc-by-nc-4.0 | tannedbum | 10 | 1.8k |
| synk/L3-8B-Stheno-v3.2-MLX | quantized | MLX | 8.03B | M | cc-by-nc-4.0 | synk | 1 | 2.1k |
| solidrust/L3-8B-Stheno-v3.2-AWQ | quantized | AWQ | 8.03B | M | — | solidrust | 2 | 1.9k |
| QuantFactory/L3-Rhaenys-8B-GGUF | merge | GGUF | — | — | cc-by-nc-4.0 | QuantFactory | 4 | 1.8k |
| tannedbum/L3-Rhaenys-8B-GGUF | merge | GGUF | — | — | cc-by-nc-4.0 | tannedbum | 6 | 1.5k |
| v000000/L3-8B-Stheno-v3.2-abliterated | merge | — | 8.03B | M | — | v000000 | 10 | 1.3k |
| Den4iiks/L3-8B-Stheno-v3.2-GGUF-IQ-Imatrix | quantized | GGUF | — | — | cc-by-nc-4.0 | Den4iiks | 0 | 1.6k |
| Mawdistical/Squelching-Fantasies-8B | finetune | — | 8.03B | M | cc-by-nd-4.0 | Mawdistical | 7 | 1.2k |
| O-TT/L3-8B-Stheno-v3.2-Q8_0-GGUF | quantized | Q8_0 | — | — | cc-by-nc-4.0 | O-TT | 0 | 1.5k |
| DreadPoor/Aspire-8B-model_stock | merge | — | 8.03B | M | cc-by-nc-4.0 | DreadPoor | 7 | 1k |
| inflatebot/L3-8B-Helium3-baseStheno | merge | — | 8.03B | M | — | inflatebot | 0 | 1.3k |
Repository changelog
14 commits
Repository changelog 14 commits
| Date | Commit | SHA |
|---|---|---|
| 2024-06-07 | Update README.md | 4bb828f |
| 2024-06-06 | Update README.md | 93da90e |
| 2024-06-06 | Update README.md | 9e223b2 |
| 2024-06-05 | Update README.md | 78c431f |
| 2024-06-05 | Update README.md | f6a3830 |
| 2024-06-05 | Rename 170c547969d6d0c2522f3a91181008eb.png to Stheno.png | 8fccb8f |
| 2024-06-05 | Upload 170c547969d6d0c2522f3a91181008eb.png | 7307cb7 |
| 2024-06-05 | Delete .ipynb_checkpoints | cd1c715 |
| 2024-06-05 | Update README.md | ff0d22e |
| 2024-06-05 | Update README.md | 381491e |
| 2024-06-05 | Update README.md | 55010b9 |
| 2024-06-05 | Create README.md | 1520110 |
| 2024-06-05 | Upload folder using huggingface_hub | 2b2b409 |
| 2024-06-05 | initial commit | 26a588a |
All providers
2 providers · 2 endpoints
Where to run 2 providers · 2 endpoints
Hosted endpoints observed for L3-8B-Stheno-v3.2, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting. Cheapest observed input price: $0.05/M.
| Provider | Model ref | Context | Quant | $/M in | $/M out | tok/s | TTFT | Tools | Status | Observed |
|---|---|---|---|---|---|---|---|---|---|---|
| Novita hf | Sao10K/L3-8B-Stheno-v3.2 | 8K | — | $0.05 | $0.05 | 110.2 | 525 ms | no | live | 2026-10-25 |
| Featherless hf | Sao10K/L3-8B-Stheno-v3.2 | — | — | — | — | — | — | — | live | 2026-10-25 |
Similar models
3 models
Similar models
- Llama-3.1-8B-Instruct · 8.03B same size band (M), same task (text-generation), more providers
- Qwen3-8B · 8.19B same size band (M), same task (text-generation), more providers, permissive license
- Meta-Llama-3-8B-Instruct · 8.03B same size band (M), same task (text-generation)
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:40 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:12 UTC · Last verified 2026-10-25
Follow L3-8B-Stheno-v3.2
Watch this model
- /feeds/models/Sao10K--L3-8B-Stheno-v3.2.xml(RSS 2.0) · every change to L3-8B-Stheno-v3.2
- /feeds/families/Sao10K--L3-Stheno-v3.2.xml(RSS 2.0) · every change across the L3-Stheno-v3.2 family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.