Up to date
jinaai CC BY-NC 4.0 out 21 months ago
1.54B parameters, text in, text out, 501K tokens of context. 1 provider serves it.
Verified 2026-10-25 Rank #554 in LLMs · score 74.8 method v1 Sources 3 Downloads 30d 994 · #593 of 619 tracked
Changed this week: 2 new quantized builds on the Hub (41 in all).2026-10-24
Picked by rule from 1 tracked piece: the lab's own post, the most shared, a field report, the latest.
ReaderLM-v2 treats HTML-to-markdown conversion as a true translation process rather than a selective-copy task.
Jina released ReaderLM-v2, a 1.5B parameter model that converts HTML to markdown or JSON.
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; 41 quantized builds for local use.
Top Spaces running it: Hyphonical/MCP-Utilities ♥ 4 · Felguk/ReaderLM-v2 ♥ 2
Weights only, an estimate: leave room for the KV cache. Method
About 0.9 GB at INT4, 1.9 GB at FP8, 3.7 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 (2025-01-13), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
49 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.
ReaderLM-v2 is a 1.54B parameter multilingual language model designed to convert raw HTML into markdown or JSON.[1][2] The model uses a Qwen2ForCausalLM architecture and supports a combined input and output context window of up to 512K tokens.[3][4]
Generated 2026-10-07 15:09 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 | Qwen2ForCausalLM |
| Model type | qwen2 |
| Dtype | bfloat16 |
| Params by dtype | BF16: 1,543,714,304 |
| Size band | S |
| MoE | no |
| Context length | 512,768 (config.json max_position_embeddings) |
| Languages | multilingual |
| Papers | arXiv:2503.01151 |
| License | CC BY-NC 4.0 (cc-by-nc-4.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · Qwen2ForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Qwen2ForCausalLM · 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 ReaderLM-v2: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 3.7 GB · Edge (<4 GB) |
|---|---|
| FP8 (1 byte/param) | 1.9 GB · Edge (<4 GB) |
| INT4 (0.5 bytes/param) | 0.9 GB · Edge (<4 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
quantized: 41finetune: 4merge: 4
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| mradermacher/ReaderLM-v2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | mradermacher | 11 | 10.8k |
| mradermacher/ReaderLM-v2-i1-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | mradermacher | 4 | 8.2k |
| mlx-community/jinaai-ReaderLM-v2 | quantized | — | 1.54B | S | mit | mlx-community | 25 | 5.3k |
| yasserrmd/ReaderLM-v2-gguf | quantized | gguf | — | — | apache-2.0 | yasserrmd | 0 | 2.7k |
| DevQuasar-7/jinaai.ReaderLM-v2-GGUF | quantized | GGUF | — | — | — | DevQuasar-7 | 0 | 2k |
| matrixportalx/ReaderLM-v2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | matrixportalx | 1 | 1.6k |
| Svngoku/ReaderLM-v2-Q8_0-GGUF | quantized | Q8_0 | — | — | cc-by-nc-4.0 | Svngoku | 3 | 931 |
| DhruvalLabs/ReaderLM-v2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | DhruvalLabs | 0 | 984 |
| cenkierpiotr/WEBscrapePL-1.5b | finetune | GGUF | 1.54B | S | cc-by-nc-4.0 | cenkierpiotr | 0 | 954 |
| ysn-rfd/ReaderLM-v2-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | ysn-rfd | 0 | 723 |
| kunal732/QwenReader | merge | — | 1.54B | S | — | kunal732 | 0 | 708 |
| liodon-ai/ReaderLM-v2-imatrix-GGUF | quantized | imatrix | — | — | other | liodon-ai | 0 | 673 |
| Soulfate24/ReaderLM-v2-Paretrix | quantized | GGUF | — | — | cc-by-nc-4.0 | Soulfate24 | 2 | 426 |
| fdgsgqaegt/ReaderLM-v2-Q4_K_M-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | fdgsgqaegt | 3 | 345 |
| roleplaiapp/ReaderLM-v2-Q4_K_M-GGUF | quantized | GGUF | — | — | — | roleplaiapp | 1 | 381 |
| roleplaiapp/ReaderLM-v2-Q4_0-GGUF | quantized | Q4_0 | — | — | — | roleplaiapp | 1 | 375 |
| alexgusevski/ReaderLM-v2-q4-mlx | quantized | mlx | 1.54B | S | cc-by-nc-4.0 | alexgusevski | 2 | 308 |
| roleplaiapp/ReaderLM-v2-Q8_0-GGUF | quantized | Q8_0 | — | — | — | roleplaiapp | 0 | 355 |
| alexgusevski/ReaderLM-v2-mlx | finetune | mlx | 1.54B | S | cc-by-nc-4.0 | alexgusevski | 1 | 285 |
| roleplaiapp/ReaderLM-v2-Q2_K-GGUF | quantized | Q2_K | — | — | — | roleplaiapp | 0 | 323 |
| adriabama06/ReaderLM-v2-AWQ | quantized | AWQ | 1.54B | S | cc-by-nc-4.0 | adriabama06 | 1 | 223 |
| adriabama06/ReaderLM-v2-GPTQ-8bit | quantized | GPTQ | — | — | cc-by-nc-4.0 | adriabama06 | 1 | 220 |
| roleplaiapp/ReaderLM-v2-Q5_0-GGUF | quantized | Q5_0 | — | — | — | roleplaiapp | 0 | 250 |
| roleplaiapp/ReaderLM-v2-Q5_K_M-GGUF | quantized | GGUF | — | — | — | roleplaiapp | 0 | 250 |
| roleplaiapp/ReaderLM-v2-Q5_K_S-GGUF | quantized | GGUF | — | — | — | roleplaiapp | 0 | 246 |
| roleplaiapp/ReaderLM-v2-Q6_K-GGUF | quantized | Q6_K | — | — | — | roleplaiapp | 0 | 231 |
| roleplaiapp/ReaderLM-v2-Q4_K_S-GGUF | quantized | GGUF | — | — | — | roleplaiapp | 0 | 218 |
| roleplaiapp/ReaderLM-v2-Q3_K_S-GGUF | quantized | GGUF | — | — | — | roleplaiapp | 0 | 214 |
| alexgusevski/ReaderLM-v2-q8-mlx | quantized | mlx | 1.54B | S | cc-by-nc-4.0 | alexgusevski | 0 | 191 |
| keisuke-miyako/ReaderLM-v2-gguf-q4_k_m | quantized | gguf | — | — | — | keisuke-miyako | 0 | 184 |
| ysn-rfd/ReaderLM-v2-Q8_0-GGUF | quantized | Q8_0 | — | — | cc-by-nc-4.0 | ysn-rfd | 0 | 177 |
| roleplaiapp/ReaderLM-v2-Q3_K_L-GGUF | quantized | GGUF | — | — | — | roleplaiapp | 0 | 170 |
| yikerman/ReaderLM-v2-Q8_0-GGUF | quantized | Q8_0 | — | — | cc-by-nc-4.0 | yikerman | 0 | 135 |
| KrzTyb/ReaderLM-v2-Q4_K_M-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | KrzTyb | 0 | 133 |
| roboalchemist/ReaderLM-v2-mlx-fp16 | finetune | mlx | 1.54B | S | cc-by-nc-4.0 | roboalchemist | 0 | 127 |
| alexgusevski/ReaderLM-v2-q3-mlx | quantized | mlx | 1.54B | S | cc-by-nc-4.0 | alexgusevski | 0 | 126 |
| Edge-Quant/ReaderLM-v2-Q4_K_M-GGUF | quantized | GGUF | — | — | cc-by-nc-4.0 | Edge-Quant | 0 | 105 |
| logxdx/ReaderLM-v2-Q8_0-GGUF | quantized | Q8_0 | — | — | cc-by-nc-4.0 | logxdx | 0 | 103 |
| kunal732/qwenreader4 | merge | — | 1.54B | S | — | kunal732 | 0 | 97 |
| alexgusevski/ReaderLM-v2-q6-mlx | quantized | mlx | 1.54B | S | cc-by-nc-4.0 | alexgusevski | 0 | 85 |
| Date | Commit | SHA |
|---|---|---|
| 2025-03-04 | add arxiv link | 1d07078 |
| 2025-02-06 | fix safetensor weights | ac3438b |
| 2025-01-22 | fix lm.head weights | 2dd358a |
| 2025-01-20 | fix error in the html-to-json example | 93b98c1 |
| 2025-01-17 | Update README.md | 352f47d |
| 2025-01-17 | Upload ONNX weights (#6) | 1c52ef1 |
| 2025-01-16 | fix: tokenizer in prompt building | 2a41ea3 |
| 2025-01-16 | Update README.md (#2) | 9cb0bd3 |
| 2025-01-15 | fix: fix the prompt in the prompt | 5c64de7 |
| 2025-01-15 | feat: reformat codes using black | c3ad4e9 |
| 2025-01-15 | Update README.md | 27de169 |
| 2025-01-15 | Update README.md | c79aa4b |
| 2025-01-15 | Update README.md | af99331 |
| 2025-01-15 | feat: update the README | 1256ad3 |
| 2025-01-15 | Update README.md | bd791d0 |
| 2025-01-15 | Update README.md | 14fbb0c |
| 2025-01-15 | Update README.md | a6c0d30 |
| 2025-01-15 | init commit | 6e1f1ac |
| 2025-01-13 | init commit | b79948b |
| 2025-01-13 | init commit | 65f8b66 |
| 2025-01-13 | init commit | fa2224f |
| 2025-01-13 | init commit | 107ed28 |
| 2025-01-13 | init commit | 0df23dc |
| 2025-01-13 | initial commit | eb684a8 |
Hosted endpoints observed for ReaderLM-v2, 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 | jinaai/ReaderLM-v2 | — | — | — | — | — | — | — | 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 21:39 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:09 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.
