Up to date
Qwen Apache 2.0 out 12 months ago
2.13B parameters, images and text in, text out, 256K tokens of context. 1 provider serves it.
Verified 2026-10-25 Rank #78 in LLMs · score 103.3 method v1 Sources 3 Downloads 30d 1.9M · #66 of 619 tracked
Picked by rule from 15 tracked pieces: the lab's own post, the most shared, a field report, the latest.
OpenAI released gpt-oss-120b and gpt-oss-20b, their first open-weight models since GPT-2 in 2019.
The 3 picks above, plus these 12. 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; heavily run locally (97 quantized builds); a big base for builders (400 fine-tunes, adapters and merges).
Top Spaces running it: prithivMLmods/Qwen3-VL-Outpost ♥ 208 · embedl/hfviewer ♥ 181 · M3st3rJ4k3l/FLUX.2-Klein-Multi-LoRA-v2 ♥ 133 · artificialguybr/qwen-vl ♥ 126
Weights only, an estimate: leave room for the KV cache. Method
About 1.3 GB at INT4, 2.6 GB at FP8, 5.1 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 (2025-10-19), amber triangles are 5 dated articles (each links to it), grey ticks are repo commits.
497 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.
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 | image-text-to-text |
|---|---|
| Library | transformers |
| Architectures | Qwen3VLForConditionalGeneration |
| Model type | qwen3_vl |
| Dtype | bfloat16 |
| Params by dtype | BF16: 2,127,532,032 |
| Size band | M |
| MoE | no |
| Context length | 262,144 (config.json max_position_embeddings) |
| Papers | arXiv:2505.09388, arXiv:2502.13923, arXiv:2409.12191, arXiv:2308.12966 |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · Qwen3VLForConditionalGeneration · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Qwen3VLForConditionalGeneration · 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 Qwen3-VL-2B-Instruct: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 5.1 GB · Single GPU (<24 GB) |
|---|---|
| FP8 (1 byte/param) | 2.6 GB · Edge (<4 GB) |
| INT4 (0.5 bytes/param) | 1.3 GB · Edge (<4 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 284quantized: 97adapter: 116
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| Qwen/Qwen3-VL-Reranker-2B | finetune | — | 2.13B | M | apache-2.0 | Qwen | 237 | 4.6M |
| unsloth/Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | apache-2.0 | unsloth | 47 | 1.9M |
| Qwen/Qwen3-VL-2B-Instruct-FP8 | quantized | -FP8 | 2.44B | M | apache-2.0 | Qwen | 54 | 1.4M |
| cyankiwi/Qwen3-VL-2B-Instruct-AWQ-4bit | quantized | AWQ | 2.13B | M | apache-2.0 | cyankiwi | 1 | 1M |
| unsloth/Qwen3-VL-2B-Instruct-unsloth-bnb-4bit | quantized | -bnb | 2.17B | M | apache-2.0 | unsloth | 8 | 880.8k |
| typhoon-ai/typhoon-ocr1.5-2b | finetune | — | 2.13B | M | apache-2.0 | typhoon-ai | 30 | 854.8k |
| Qwen/Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | apache-2.0 | Qwen | 72 | 688.5k |
| ADSKAILab/Zero-To-CAD-Qwen3-VL-2B | finetune | — | 2.13B | M | apache-2.0 | ADSKAILab | 83 | 129.8k |
| unsloth/Qwen3-VL-2B-Instruct | finetune | — | 2.13B | M | apache-2.0 | unsloth | 9 | 117.9k |
| FireRedTeam/FireRed-OCR | finetune | — | 2.13B | M | apache-2.0 | FireRedTeam | 167 | 82.8k |
| bartowski/Qwen_Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | — | bartowski | 2 | 50.1k |
| AXERA-TECH/Qwen3-VL-8B-Instruct-GPTQ-Int4 | finetune | GPTQ | — | — | mit | AXERA-TECH | 1 | 47.6k |
| ggml-org/Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | — | ggml-org | 8 | 31.4k |
| unsloth/Qwen3-VL-2B-Instruct-1M-GGUF | quantized | GGUF | — | — | apache-2.0 | unsloth | 8 | 26.2k |
| unsloth/Qwen3-VL-2B-Instruct-bnb-4bit | quantized | -bnb | 2.17B | M | apache-2.0 | unsloth | 4 | 24.6k |
| lmstudio-community/Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | apache-2.0 | lmstudio-community | 1 | 21.5k |
| furiosa-ai/Qwen3-VL-2B-Instruct | finetune | — | 2.13B | M | apache-2.0 | furiosa-ai | 0 | 21.3k |
| huihui-ai/Huihui-Qwen3-VL-2B-Instruct-abliterated | finetune | — | 2.13B | M | apache-2.0 | huihui-ai | 22 | 17k |
| ce-amtic/ProcVLM-2B | finetune | — | 2.44B | M | cc-by-4.0 | ce-amtic | 0 | 17.7k |
| mobilint/Qwen3-VL-2B-Instruct | quantized | — | 314M | S | apache-2.0 | mobilint | 0 | 15.8k |
| lmstudio-community/Qwen3-VL-2B-Instruct-MLX-8bit | quantized | MLX | 2.13B | M | apache-2.0 | lmstudio-community | 4 | 8.9k |
| mradermacher/Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | apache-2.0 | mradermacher | 1 | 8.6k |
| prithivMLmods/Qwen3-VL-2B-Instruct-abliterated-v1 | finetune | — | 2.13B | M | apache-2.0 | prithivMLmods | 3 | 8.3k |
| lmstudio-community/Qwen3-VL-2B-Instruct-MLX-bf16 | finetune | MLX | 2.13B | M | apache-2.0 | lmstudio-community | 2 | 7.4k |
| mradermacher/Qwen3-VL-2B-Instruct-i1-GGUF | quantized | GGUF | — | — | apache-2.0 | mradermacher | 0 | 6.4k |
| mohajesmaeili/Qwen3-VL-2B-Persian-Arabic-Ocr-v1.0 | finetune | — | 2.13B | M | apache-2.0 | mohajesmaeili | 22 | 4.2k |
| phxember/Uni-MuMER-Qwen3-VL-2B | finetune | — | 2.13B | M | apache-2.0 | phxember | 0 | 5.1k |
| CodeGoat24/UnifiedReward-2.0-qwen3vl-2b | finetune | — | 2.13B | M | mit | CodeGoat24 | 2 | 4.6k |
| tatsuyaaaaaaa/Qwen3-VL-2B-Instruct-gguf | quantized | gguf | — | — | apache-2.0 | tatsuyaaaaaaa | 0 | 4.5k |
| rokotmidi/rokot-omr-2b | quantized | GGUF | 2.13B | M | cc-by-nc-4.0 | rokotmidi | 3 | 4.3k |
| arthurcollet/Qwen3-VL-Embedding-2B-mlx | finetune | mlx | 2.13B | M | apache-2.0 | arthurcollet | 2 | 4.3k |
| inclusionAI/SingGuard-2b | finetune | — | 2.13B | M | apache-2.0 | inclusionAI | 15 | 3.3k |
| geolazagr/greko | quantized | GGUF | — | — | apache-2.0 | geolazagr | 2 | 3.9k |
| SberRoboticsCenter/Qwen3-VL-2B-Instruct-action | finetune | — | 2.13B | M | — | SberRoboticsCenter | 0 | 3.8k |
| mlx-community/Qwen3-VL-Embedding-2B-4bit | finetune | -4bit | 2.13B | M | apache-2.0 | mlx-community | 2 | 3.4k |
| InternRobotics/InternVLA-A1-3B | finetune | — | 3.22B | M | cc-by-nc-sa-4.0 | InternRobotics | 40 | 1.5k |
| prithivMLmods/Qwen3-VL-2B-Instruct-GGUF | quantized | GGUF | — | — | apache-2.0 | prithivMLmods | 1 | 3.2k |
| mlx-community/Qwen3-VL-Embedding-2B-8bit | finetune | -8bit | 2.13B | M | apache-2.0 | mlx-community | 0 | 3.3k |
| MCG-NJU/TimeLens2-2B | finetune | — | 2.44B | M | apache-2.0 | MCG-NJU | 8 | 2.4k |
| mlx-community/Qwen3-VL-Embedding-2B-mxfp8 | finetune | — | 2.13B | M | apache-2.0 | mlx-community | 1 | 2.6k |
Hosted endpoints observed for Qwen3-VL-2B-Instruct, 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 | Qwen/Qwen3-VL-2B-Instruct | — | — | — | — | — | — | — | 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 21:33 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-07 15:00 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.
