All models
Alibaba (Qwen)
Qwen3.5-9B fastruns locally
alibaba/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...
Input / 1M
$0.17
USD per million tokens
Output / 1M
$0.25
USD per million tokens
Est. cost / token
$0.00000019
$0.19 per 1M, 3:1 mix
Context window
262K
max output 236K
Major features
VisionOpen weightsReasoningJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding3
- reasoning3
- math3
- summarization4
- extraction4
- classification4
- chat3
- creative3
- translation4
- vision3
- long context3
- agentic3
Run it locally
Will it run on my machine? →| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 6 GB |
| Q3_K_M | fair | 6.7 GB |
| Q4_K_Mrecommended | good | 8 GB |
| Q5_K_M | very good | 9 GB |
| Q6_K | excellent | 10.1 GB |
| Q8_0 | near-lossless | 12.6 GB |
| F16 | lossless | 22.1 GB |
Estimates.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.17 | $0.25 | — | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.17
- Output
- $0.25
- Blended (3:1)
- $0.19
Details
- Company
- Alibaba (Qwen)
- Version
- 3.5
- Family
- Qwen
- Released
- 2026-03-10
- Parameters
- 9.7B
- Max output
- 236K
- Open weights
- Yes
- Inputs
- image, text, video
- Last synced
- 25 min ago