Qwen3.5-9B
Qwenhuggingface.co/Qwen/Qwen3.5-9B
Qwen3.5-9B is a multimodal foundation model from the Qwen3.5 family, designed to deliver strong reasoning, coding, and visual understanding in an efficient 9B-parameter architecture. It uses a unified vision-language design...
Updated 54 min ago · first seen 11 Sept 2026
model_01M294WW3ER0NCZXFMCX80TXGW
- Parameters
- 9.65B
- T2 · 6 h ago
- Context
- 262.1K tokens
- T2 · 6 h ago
- Released
- 27 Feb 2026
- T2 · 54 min ago
- License
- apache-2.0
- T2 · 5 h ago
Specification
- Release date
- 27 Feb 2026
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 54 min agomedium
- Openness
- open-weights
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 6 h agomedium
- License
- apache-2.0
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Architecture
- Qwen3_5ForConditionalGeneration
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Parameters
- 9.65B
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 6 h agomedium
- Context window
- 262.1K tokens
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Max output
- 235.9K tokens
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Modalities
- image, text, video
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Input modalities
- image, text, video
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Output modalities
- text
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Tokenizer
- Qwen3
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Base model
- Qwen/Qwen3.5-9B-Base
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- File size
- 19.3 GB
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Hugging Face repo
- Qwen/Qwen3.5-9B
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Pipeline tag
- image-text-to-text
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 6 h agomedium
- Model card
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 6 h agomedium
- Downloads
- 10,588,820
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 54 min agomedium
- Likes
- 1,924
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 54 min agomedium
- Gated
- No
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Hf inference providers
- deepinfra, featherless-ai, ovhcloud, together
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Last modified
- 2026-03-02T00:51:43+00:00
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 6 h agomedium
- Library name
- transformers
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- License url
- https://huggingface.co/Qwen/Qwen3.5-9B/blob/main/LICENSE
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Aa median output tokens per second
- 87.3
Source:Artificial AnalysisT2observed 4 h agomedium
- Downloads all time
- 60,507,487
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 54 min agomedium
- Model type
- qwen3_5
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Openrouter id
- qwen/qwen3.5-9b
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Supported parameters
- frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Tags
- transformers, safetensors, qwen3_5, image-text-to-text
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
- Weights dtype
- BF16, F32
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium
Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →
Provenance
Attributed facts
35
Source tiers
T235
Freshest observation
54 min ago
Conflicts
None
Modalities
- Modalities
- imagetextvideo
- Input
- imagetextvideo
- Output
- text
Capabilities
Tool calling
Yes
OpenRouter public model & pricing listing · T2
Structured output
Yes
OpenRouter public model & pricing listing · T2
Reasoning
Yes
OpenRouter public model & pricing listing · T2
Vision
Yes
OpenRouter public model & pricing listing · T2
Audio
Unavailable
Fine-tuning available
Unavailable
- Context window
- 262.1K tokens
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Max output
- 235.9K tokens
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
- Tokenizer
- Qwen3
Source:OpenRouter public model & pricing listingT2observed 6 h agomedium
| Benchmark | Score | Metric | Config | Evaluated | Source |
|---|---|---|---|---|---|
| τ²-bench | 86.84% | pass^1 | aa_slug=qwen3-5-9b · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
| Artificial Analysis Intelligence Index | 13.65 | index | aa_slug=qwen3-5-9b · version=4.3 · estimated=true | — | artificialanalysis.aiT2 |
| GPQA | 80.61% | accuracy | aa_slug=qwen3-5-9b · variant=GPQA Diamond · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
| Humanity's Last Exam | 14.92% | accuracy | aa_slug=qwen3-5-9b · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
| IFBench | 66.73% | accuracy | aa_slug=qwen3-5-9b · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
| MMMU-Pro | 69.25% | accuracy | aa_slug=qwen3-5-9b · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
| Terminal-Bench | 29.21% | accuracy | aa_slug=qwen3-5-9b · variant=v2.1 · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
| Terminal-Bench | 24.24% | accuracy | aa_slug=qwen3-5-9b · variant=hard · evaluator=Artificial Analysis · index_version=4.3 | — | artificialanalysis.aiT2 |
Scores are reported as published, with their evaluation configuration (harness, prompting, judge). Results with different configs are not directly comparable — see methodology.
Current prices
| Provider | Input / 1M | Output / 1M | Cached in | Batch in / out | Context | Observed | Source |
|---|---|---|---|---|---|---|---|
| OpenRouterqwen/qwen3.5-9b | $0.10 | $0.15 | — | — | 262.1K | 2 h ago | openrouter.aiT2 |
| OpenRouterqwen/qwen3.5-9b:batch | $0.17 | $0.25 | — | — | 262.1K | 2 h ago | openrouter.aiT2 |
| Together AIqwen3-5-9b | $0.17 | $0.25 | — | — | — | 4 h ago | together.aiT1 |
USD per 1M tokens as published by each provider (USD). Rows are append-only: every change is kept in the history below.
Available through 2
Price history
Input price · USD / 1M tokens
Output price · USD / 1M tokens
- OpenRouter
- Together AI
Memory need = bytes per parameter (4-bit ≈ 0.5 × 1.15 overhead, 8-bit 1.0, fp16 2.0) + a KV-cache allowance. Not a measurement.
| Hardware | Quantization | Memory | Est. need | Fits |
|---|---|---|---|---|
| NVIDIA DGX B200 | 4bit | 1,440 GB | 6.0 GB | Yes |
| Apple M3 Ultra | 4bit | — | 6.0 GB | Yes |
| Mac Studio (Apple M5 Ultra) | 4bit | — | 6.0 GB | Yes |
| AMD Instinct MI325X | 4bit | 256 GB | 6.0 GB | Yes |
| AMD Instinct MI300X | 4bit | 192 GB | 6.0 GB | Yes |
| Apple M2 Ultra | 4bit | — | 6.0 GB | Yes |
| NVIDIA B200 | 4bit | 180 GB | 6.0 GB | Yes |
| NVIDIA H200 | 4bit | 141 GB | 6.0 GB | Yes |
| NVIDIA H200 NVL | 4bit | 141 GB | 6.0 GB | Yes |
| Apple M1 Ultra | 4bit | — | 6.0 GB | Yes |
| Apple M3 Max | 4bit | — | 6.0 GB | Yes |
| Apple M4 Max | 4bit | — | 6.0 GB | Yes |
| Mac Studio (Apple M5 Max) | 4bit | — | 6.0 GB | Yes |
| MacBook Pro (Apple M5 Max) | 4bit | — | 6.0 GB | Yes |
| NVIDIA DGX Spark | 4bit | 128 GB | 6.0 GB | Yes |
| Apple M2 Max | 4bit | — | 6.0 GB | Yes |
| NVIDIA H100 NVL | 4bit | 94 GB | 6.0 GB | Yes |
| NVIDIA A100 80GB | 4bit | 80 GB | 6.0 GB | Yes |
| NVIDIA H100 SXM | 4bit | 80 GB | 6.0 GB | Yes |
| Apple M1 Max | 4bit | — | 6.0 GB | Yes |
| Apple M4 Pro | 4bit | — | 6.0 GB | Yes |
| Mac mini (Apple M5 Pro) | 4bit | — | 6.0 GB | Yes |
| MacBook Pro (Apple M5 Pro) | 4bit | — | 6.0 GB | Yes |
| Apple M3 Pro | 4bit | — | 6.0 GB | Yes |
| Apple M1 Pro | 4bit | — | 6.0 GB | Yes |
| Apple M2 Pro | 4bit | — | 6.0 GB | Yes |
| Apple M4 | 4bit | — | 6.0 GB | Yes |
| iMac (Apple M4) | 4bit | — | 6.0 GB | Yes |
| Mac mini (Apple M6) | 4bit | — | 6.0 GB | Yes |
| MacBook Air (Apple M5) | 4bit | — | 6.0 GB | Yes |
| MacBook Pro (Apple M5) | 4bit | — | 6.0 GB | Yes |
| NVIDIA GeForce RTX 5090 | 4bit | 32 GB | 6.0 GB | Yes |
| Apple M2 | 4bit | — | 6.0 GB | Yes |
| Apple M3 | 4bit | — | 6.0 GB | Yes |
| NVIDIA GeForce RTX 3090 | 4bit | 24 GB | 6.0 GB | Yes |
| NVIDIA GeForce RTX 4090 | 4bit | 24 GB | 6.0 GB | Yes |
| Apple M1 | 4bit | — | 6.0 GB | Yes |
Ancestors 1
- Qwen3.5-9B-BaseQwen
This model
Qwen3.5-9B
9.65B params
Quantizations 2
Descendants 0
None recorded.
Papers 0
No papers linked yet.
Repositories 0
No repositories linked yet.
As of
Rewind the record: see this entity's attributes exactly as AI Atlas knew them on a given day.
Claim history · Parameters
Parametersparameter_count1
| Value | Valid from → to | Status | Source | Confidence | Extractor |
|---|---|---|---|---|---|
| 9.65B | → current | current | Hugging Face Hub (public pages, model cards, papers)T2 | medium | deterministic |
Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →
Qwen3.5-9B scores 14.92% on Humanity's Last Exam
artificial_analysisQwen3.5-9B scores 13.65 on Artificial Analysis Intelligence Index
artificial_analysisTogether AI lists Qwen3.5-9B at $0.17 in / $0.25 out per 1M tokens
together_pricingQwen3.5-9B: release date changed from 2026-02-27 to 2026-03-10
Release date27 Feb 2026→10 Mar 2026openrouterQwen3.5-9B: release date changed from 2026-03-10 to 2026-02-27
Release date10 Mar 2026→27 Feb 2026huggingfaceOpenRouter lists Qwen3.5-9B at $0.17 in / $0.25 out per 1M tokens
openrouterOpenRouter lists Qwen3.5-9B at $0.1 in / $0.15 out per 1M tokens
openrouter
| Source | Document | Type | Tier | Last observed | Snapshots |
|---|---|---|---|---|---|
| Together AI — pricing | together.ai/pricing | pricing | T1· Official | 4 h ago | 1 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/unsloth/Qwen3.5-9B-GGUF | model_page | T2· Quality secondary | 48 min ago | 3 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/Qwen/Qwen3.5-9B | model_page | T2· Quality secondary | 54 min ago | 4 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/Intel/Qwen3.5-9B-int4-AutoRound | model_page | T2· Quality secondary | 1 h ago | 4 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/models?author=Qwen&p=0&sort=downloads | listing | T2· Quality secondary | 1 h ago | 4 |
| OpenRouter public model & pricing listing | openrouter.ai/api/v1/models | catalogue | T2· Quality secondary | 2 h ago | 4 |
| Hugging Face Hub (public pages, model cards, papers) | huggingface.co/Qwen/Qwen3.5-9B/raw/main/README.md | model_card | T2· Quality secondary | 3 h ago | 1 |
| Artificial Analysis | artificialanalysis.ai/leaderboards/models | leaderboard | T2· Quality secondary | 4 h ago | 1 |
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.