Up to date
microsoft Research / non-commercial (microsoft-research-license) out 3 years ago
Text in, text out, 4K tokens of context. 1 provider serves it.
Verified 2026-10-25 Rank #540 in LLMs · score 75.3 method v1 Sources 3 Downloads 30d 987 · #594 of 619 tracked
Picked by rule from 5 tracked pieces: the lab's own post, the most shared, a field report, the latest.
Orca 2 teaches small language models to reason by training them on diverse solution strategies like step-by-step and recall-then-generate rather than just imitating larger models.
Orca-2-13B is a model available in 7B and 13B parameter versions, fine-tuned from the LLAMA 2 base models to excel in reasoning and math tasks.
The 2 picks above, plus these 3. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.
| # | Cheapest live row | In /1M | Out /1M | Context | Quant | Seen |
|---|---|---|---|---|---|---|
| 1 | Featherless | — | — | — | — | Oct 25 |
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 1 API provider; 11 quantized builds for local use.
Top Spaces running it: ari9dam/Orca-2-13B ♥ 34 · CATIE-AQ/Guide_Evaluation_LLM ♥ 5 · tonic-ai/Orca213B ♥ 3 · li-qing/FIRE ♥ 3
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2023-11-14), amber triangles are 5 dated articles (each links to it), grey ticks are repo commits.
21 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
Orca 2 is a 13B parameter LlamaForCausalLM model designed for research tasks like reasoning, math problem solving, and text summarization.[1][2][3]
Generated 2026-10-07 15:13 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.
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.
Listed in LLMs
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | LlamaForCausalLM |
| Model type | llama |
| Dtype | float32 |
| MoE | no |
| Context length | 4,096 (config.json max_position_embeddings) |
| Papers | arXiv:2311.11045 |
| License | Research / non-commercial (microsoft-research-license) (other) |
| Hub snapshot | 2026-10-25 |
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.
finetune: 5quantized: 11merge: 2adapter: 3
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| Locutusque/Orca-2-13b-SFT-v4 | finetune | — | 13.0B | L | other | Locutusque | 5 | 213.3k |
| Locutusque/Orca-2-13b-SFT_v5 | finetune | — | 13.0B | L | other | Locutusque | 1 | 194.3k |
| Locutusque/Orca-2-13b-SFT-v6 | finetune | — | 13.0B | L | other | Locutusque | 10 | 188.4k |
| TheBloke/Orca-2-13B-AWQ | quantized | AWQ | 13.0B | L | other | TheBloke | 7 | 75.4k |
| TheBloke/Orca-2-13B-GGUF | quantized | GGUF | — | — | other | TheBloke | 66 | 53.4k |
| tuantran1632001/Psyfighter2-Orca2-13B-ties | merge | — | 13.0B | L | other | tuantran1632001 | 2 | 20.9k |
| second-state/Orca-2-13B-GGUF | quantized | GGUF | — | — | other | second-state | 7 | 20.5k |
| TheBloke/Orca-2-13B-GPTQ | quantized | GPTQ | 13.0B | L | other | TheBloke | 27 | 15.8k |
| mradermacher/Orca-2-13b-i1-GGUF | quantized | GGUF | — | — | other | mradermacher | 0 | 8.4k |
| mradermacher/Orca-2-13b-GGUF | quantized | GGUF | — | — | other | mradermacher | 0 | 7.2k |
| TachyHealth/Thealth-Microsoft-Orca-2-13b | adapter | — | 13.0B | L | ms-pl | TachyHealth | 1 | 470 |
| ricecake/Orca-2-13B-Pygmalion-LoRA | quantized | — | — | — | other | ricecake | 2 | 307 |
| athirdpath/Orca-2-13b-Alpaca-Uncensored-LORA | quantized | — | — | — | other | athirdpath | 2 | 266 |
| Chat-Error/Testing_orca | quantized | — | — | — | other | Chat-Error | 1 | 232 |
| mgoin/Orca-2-13b-pruned50-quant-ds | quantized | -quant | — | — | — | mgoin | 0 | 269 |
| ssaryssane/ssarry-truthful-13B-slerp | merge | — | 10.5B | L | — | ssaryssane | 0 | 227 |
| itlwas/Orca-2-13b-Q4_K_M-GGUF | quantized | GGUF | — | — | other | itlwas | 0 | 219 |
| ricecake/Orca-2-13B-Pyg-and-Bluemoon | adapter | — | — | — | — | ricecake | 3 | 36 |
| bartowski/Orca-2-13B-no_robots-exl2 | finetune | exl2 | — | — | other | bartowski | 1 | 32 |
| gianlucar/orca2_finetuned_test_contenzioso | adapter | — | — | — | — | gianlucar | 0 | 82 |
| noneUsername/Orca-2-13b-W8A8-Dynamic-Per-Token | finetune | -W8A8 | 13.0B | L | — | noneUsername | 0 | 52 |
Hosted endpoints observed for Orca-2-13b, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting.
| Provider | Model ref | Context | Quant | $/M in | $/M out | tok/s | TTFT | Tools | Status | Observed |
|---|---|---|---|---|---|---|---|---|---|---|
| Featherless hf | microsoft/Orca-2-13b | — | — | — | — | — | — | — | live | 2026-10-25 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:42 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:13 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
