Up to date
NousResearch Not declared out 3 years ago
6.74B parameters, text in, text out, 4K tokens of context. 1 provider serves it.
Verified 2026-10-25 Rank #521 in LLMs · score 75.7 method v1 Sources 3 Downloads 30d 87k · #329 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; 11 quantized builds for local use; a big base for builders (225 fine-tunes, adapters and merges).
Top Spaces running it: Yeyito/llm_contamination_detector ♥ 27 · Ligeng-Zhu/VILA ♥ 4 · bfshi/VILA-HD-demo ♥ 4 · Ailaysa-AI/Asai-Tamil-Tokenizer ♥ 4
Weights only, an estimate: leave room for the KV cache. Method
About 4.0 GB at INT4, 8.1 GB at FP8, 16 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.
236 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 | 29.3 | Llama 2 7B | model author | 2026-10-07 |
| author-reported BBH · Reason through hard problems | 32.6 | Llama 2 7B | model author | 2026-10-07 |
| author-reported Code | 16.8 | Llama 2 7B | model author | 2026-10-07 |
| author-reported Commonsense Reasoning | 63.9 | Llama 2 7B | model author | 2026-10-07 |
| author-reported MATH-500 · Reason through hard problems | 14.6 | Llama 2 7B | model author | 2026-10-07 |
| author-reported MMLU · Reason through hard problems | 45.3 | Llama 2 7B | model author | 2026-10-07 |
| author-reported Reading Comprehension | 61.3 | Llama 2 7B | model author | 2026-10-07 |
| author-reported Toxigen | 21.25 | Llama 2 7B | model author | 2026-10-07 |
| author-reported Toxigen | 0 | Llama-2-Chat 7B | model author | 2026-10-07 |
| author-reported TruthfulQA · Chat and follow instructions | 33.29 | Llama 2 7B | model author | 2026-10-07 |
| author-reported TruthfulQA · Chat and follow instructions | 57.04 | Llama-2-Chat 7B | model author | 2026-10-07 |
| author-reported World Knowledge | 48.9 | Llama 2 7B | model author | 2026-10-07 |
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: 2,048 · F16: 6,738,415,616 |
| Size band | M |
| 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-2-7b-hf: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 16 GB · Single GPU (<24 GB) |
|---|---|
| FP8 (1 byte/param) | 8.1 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 4.0 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 52quantized: 11adapter: 139merge: 34
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| teknium/OpenHermes-7B | finetune | — | — | — | mit | teknium | 14 | 57.5k |
| mradermacher/Llama-2-7b-hf-i1-GGUF | quantized | GGUF | — | — | — | mradermacher | 0 | 5.4k |
| SeanLee97/bellm-llama-7b-nli | adapter | — | — | — | — | SeanLee97 | 2 | 2.9k |
| mradermacher/Llama-2-7b-hf-GGUF | quantized | GGUF | — | — | — | mradermacher | 0 | 2.9k |
| nateraw/llama-2-7b-english-to-hinglish | adapter | — | — | — | apache-2.0 | nateraw | 11 | 1.4k |
| igorktech/RuBit-LLama-63M | finetune | — | 63.3M | S | — | igorktech | 1 | 1.1k |
| ekaterinatao/translation_llm | adapter | — | — | — | — | ekaterinatao | 0 | 971 |
| Damarinm26/fingpt-analista-tfm | adapter | GGUF | — | — | — | Damarinm26 | 0 | 751 |
| arcee-ai/Patent-Instruct-Internlm2-7B-Ties | merge | — | 6.74B | M | — | arcee-ai | 0 | 730 |
| arcee-ai/Patent-Base-InternLM2-7B-Ties | merge | — | 6.74B | M | — | arcee-ai | 0 | 711 |
| arcee-ai/Patent-Base-Orca-2-7B-Ties | merge | — | 6.74B | M | — | arcee-ai | 0 | 697 |
| ueihieu/llama2-7b-lora-genwiki | adapter | — | — | — | — | ueihieu | 0 | 561 |
| WhereIsAI/billm-llama-7b-conll03-ner | adapter | — | — | — | mit | WhereIsAI | 1 | 457 |
| ueihieu/llama2-7b-lora-scierc | adapter | — | — | — | — | ueihieu | 0 | 491 |
| bratao/llama7b-finetuned-openie-lora | finetune | — | — | — | — | bratao | 1 | 438 |
| ueihieu/llama2-7b-lora-rebel | adapter | — | — | — | — | ueihieu | 0 | 479 |
| ueihieu/llama-2-7b-genwiki-context | adapter | — | — | — | — | ueihieu | 0 | 434 |
| Around6827/AdvertLlama-7b | adapter | — | — | — | — | Around6827 | 0 | 413 |
| BiniyamAjaw/llama-2-7b-finetuned-adapters | adapter | — | — | — | mit | BiniyamAjaw | 4 | 206 |
| Llama-Math/llama-polya-v2 | finetune | — | — | — | — | Llama-Math | 0 | 331 |
| openaccess-ai-collective/llama-7b-llava-1_5-pretrained-projector | finetune | — | — | — | — | openaccess-ai-collective | 0 | 327 |
| zeyadusf/llama2-miniguanaco | adapter | — | — | — | apache-2.0 | zeyadusf | 0 | 324 |
| sasvata/Llama2-7b-MOM-Summary-Finetuned | finetune | — | — | — | llama2 | sasvata | 0 | 301 |
| daveydhruti/Llama2-based-NIDS | adapter | — | — | — | — | daveydhruti | 0 | 291 |
| Pomilon/LEMA-llama-2-7b | finetune | — | 6.74B | M | mit | Pomilon | 1 | 216 |
| orderondemand/Bitnet-Llama-105M | finetune | — | 106M | S | — | orderondemand | 0 | 261 |
| CaioXapelaum/Orca-2-7b-Patent-Instruct-Llama-2 | merge | — | 6.74B | M | — | CaioXapelaum | 1 | 193 |
| juancopi81/llama-7-int4-orca-spanish | finetune | int4 | 6.74B | M | — | juancopi81 | 0 | 242 |
| Undi95/toxicqa-Llama2-7B-lora | adapter | — | — | — | — | Undi95 | 0 | 238 |
| cadaeic/llama2-7b-recipe-lora | quantized | — | — | — | — | cadaeic | 0 | 231 |
| wwhwwhwwh/mergekit-ties-bpthjul | merge | — | 6.74B | M | — | wwhwwhwwh | 0 | 227 |
| imkalpx/llama2-med-summarizer-gguf | adapter | gguf | — | — | llama2 | imkalpx | 2 | 123 |
| Abin7/finnish-mal | merge | — | 6.74B | M | — | Abin7 | 0 | 194 |
| Abin7/finnish-mal-spanish | merge | — | 6.74B | M | — | Abin7 | 0 | 194 |
| Abin7/6-lang | merge | — | 6.74B | M | — | Abin7 | 0 | 193 |
| girrajjangid/Llama-7B-SFT | quantized | — | — | — | — | girrajjangid | 0 | 183 |
| sijiasijia/lora-backward-1000 | adapter | — | — | — | mit | sijiasijia | 0 | 179 |
| mergekit-community/mergekit-ties-ujwvugo | merge | — | 6.74B | M | — | mergekit-community | 0 | 178 |
| zorooo/MathLlama-7b | adapter | — | 6.74B | M | — | zorooo | 0 | 172 |
| MaxZabarka/classifier-7b-v9 | quantized | — | — | — | — | MaxZabarka | 0 | 170 |
| Date | Commit | SHA |
|---|---|---|
| 2024-06-03 | Update generation_config.json (#9) | 8efe6c9 |
| 2023-08-26 | Delete added_tokens.json | dacdfcd |
| 2023-07-22 | revert model_max_length | 42c97b7 |
| 2023-07-21 | Fix extremely long model_max_length (#2) | b3ced1b |
| 2023-07-18 | Update config.json | 7ad23d9 |
| 2023-07-18 | Update README.md | 280ec9f |
| 2023-07-18 | Squashing commit | 637a748 |
Hosted endpoints observed for Llama-2-7b-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-2-7b-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 21:39 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:11 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.
