Up to date
NousResearch Not declared out 3 years ago
13.0B parameters, text in, text out, 4K tokens of context. 1 provider serves it.
Verified 2026-10-25 Rank #767 in LLMs · score 71.2 method v1 Sources 3 Downloads 30d 3.7k · #560 of 619 tracked
Picked by rule from 28 tracked pieces: the lab's own post, the most shared, a field report, the latest.
Fine-tuning open-source Llama 2 models can replace GPT-3.5 and GPT-4 for specific tasks, with a fine-tuned Llama 7B matching GPT-4 labels 95% of the time on a test set.
The 2 picks above, plus these 23. 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; 5 quantized builds for local use.
Weights only, an estimate: leave room for the KV cache. Method
About 7.8 GB at INT4, 16 GB at FP8, 31 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2023-07-18), amber triangles are 10 dated articles (each links to it), grey ticks are repo commits.
29 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.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported AGI Eval | 39.1 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported BBH · Reason through hard problems | 39.4 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported Code | 24.5 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported Commonsense Reasoning | 66.9 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported MATH-500 · Reason through hard problems | 28.7 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported MMLU · Reason through hard problems | 54.8 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported Reading Comprehension | 65.8 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported Toxigen | 26.1 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported Toxigen | 0 other | Llama-2-Chat 13B | model author | 2026-10-07 |
| author-reported TruthfulQA · Chat and follow instructions | 41.86 other | Llama 2 13B | model author | 2026-10-07 |
| author-reported TruthfulQA · Chat and follow instructions | 62.18 other | Llama-2-Chat 13B | model author | 2026-10-07 |
| author-reported World Knowledge | 55.4 other | Llama 2 13B | model author | 2026-10-07 |
This model is converted for the Hugging Face Transformers format and is intended for English natural language generation tasks.[1][2]
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 | float16 |
| Params by dtype | F32: 2,560 · F16: 13,015,864,320 |
| Size band | L |
| MoE | no |
| Context length | 4,096 (config.json max_position_embeddings) |
| Languages | en |
| License | Not declared |
| 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.
Weights-only memory for Llama-2-13b-hf: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 31 GB · Single high-end GPU (<80 GB) |
|---|---|
| FP8 (1 byte/param) | 16 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 7.8 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 11quantized: 5adapter: 11merge: 2
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| teknium/OpenHermes-13B | finetune | — | — | — | mit | teknium | 57 | 65k |
| mradermacher/Llama-2-13b-hf-i1-GGUF | quantized | GGUF | — | — | — | mradermacher | 0 | 5.1k |
| mradermacher/Llama-2-13b-hf-GGUF | quantized | GGUF | — | — | — | mradermacher | 0 | 1.7k |
| TachyHealth/Thealth-llama-13b | adapter | — | 13.0B | L | llama2 | TachyHealth | 2 | 371 |
| kubernetes-bad/Ligma-L2-13b | adapter | — | — | — | other | kubernetes-bad | 3 | 222 |
| Danielrahmai1991/model1 | quantized | GGUF | — | — | apache-2.0 | Danielrahmai1991 | 0 | 318 |
| Undi95/toxicqa-Llama2-13B | adapter | — | — | — | — | Undi95 | 6 | 0 |
| Syed-Hasan-8503/Llama-2-8B-Pruned | finetune | — | 8.89B | M | apache-2.0 | Syed-Hasan-8503 | 1 | 250 |
| LoneStriker/CollectiveCognition-v1.1-Mistral-7B-6.0bpw-h6-exl2 | finetune | 0bpw | — | — | apache-2.0 | LoneStriker | 0 | 264 |
| LoneStriker/CollectiveCognition-v1.1-Mistral-7B-8.0bpw-h6-exl2 | finetune | 0bpw | — | — | apache-2.0 | LoneStriker | 0 | 264 |
| totally-not-an-llm/EverythingLM-13b-V3-peft | adapter | — | — | — | — | totally-not-an-llm | 1 | 212 |
| Around6827/AdvertLlama-13b-chat | adapter | — | — | — | — | Around6827 | 0 | 262 |
| LoneStriker/CollectiveCognition-v1.1-Mistral-7B-3.0bpw-h6-exl2 | finetune | 0bpw | — | — | apache-2.0 | LoneStriker | 0 | 257 |
| mpasila/OpenHermes-13B-safetensors | finetune | — | 13.0B | L | mit | mpasila | 0 | 239 |
| mpasila/OpenHermes-13B-exl2-3bpw | finetune | exl2 | — | — | mit | mpasila | 0 | 232 |
| totally-not-an-llm/EverythingLM-13b-peft | adapter | — | — | — | — | totally-not-an-llm | 1 | 179 |
| KalenJosifovski/Llama-2-13b-hf-Q4_K_M-GGUF | quantized | GGUF | — | — | — | KalenJosifovski | 0 | 200 |
| LoneStriker/CollectiveCognition-v1.1-Mistral-7B-5.0bpw-h6-exl2 | finetune | 0bpw | — | — | apache-2.0 | LoneStriker | 1 | 107 |
| kobesar/fingpt-sentiment_llama2-13b_lora | adapter | — | — | — | — | kobesar | 0 | 144 |
| Undi95/toxicqa-Llama2-13B-lora | adapter | — | — | — | — | Undi95 | 2 | 38 |
| SirWumpus/FinMA | merge | — | — | — | — | SirWumpus | 0 | 92 |
| LoneStriker/CollectiveCognition-v1.1-Mistral-7B-4.0bpw-h6-exl2 | finetune | 0bpw | — | — | apache-2.0 | LoneStriker | 0 | 68 |
| fernandals/results | adapter | — | — | — | — | fernandals | 0 | 50 |
| fernandals/llama2-13b-news-gen | adapter | — | — | — | — | fernandals | 0 | 46 |
| SatoruDano/axolotl13b | quantized | — | — | — | — | SatoruDano | 0 | 43 |
| SirWumpus/FinMA2 | merge | — | — | — | — | SirWumpus | 0 | 38 |
| jirin/Llama-2-13b-fingpt | adapter | — | — | — | — | jirin | 0 | 35 |
| MR/gptj-qlora-peft | finetune | — | — | — | — | MR | 0 | 0 |
| Leooo666/llama-2-13b-recipe-qlora | finetune | — | — | — | — | Leooo666 | 0 | 0 |
Hosted endpoints observed for Llama-2-13b-hf, 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 | NousResearch/Llama-2-13b-hf | — | — | — | — | — | — | — | 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.
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:41 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.
