Qwen3-VL-8B-Instruct vs Phi-3.5-vision-instruct
Specs, VRAM requirements and download trends — updated 5 October 2026.
Qwen · A · Open weights
Params
8.8 B
Context
—
30d
5.2 M
apache-2.0 · commercial OK
microsoft · B · Open weights
Params
4.2 B
Context
131,072
30d
805 K
mit · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | Qwen3-VL-8B-Instruct | Phi-3.5-vision-instruct |
|---|---|---|
| Deployment | Open weights · also via API | Open weights |
| Parameters | 8.8 B | 4.2 B |
| Architecture | Qwen3VLForConditionalGeneration | Phi3VForCausalLM |
| Context length | 262,144 | 131,072 |
| Licence | apache-2.0 | mit |
| Commercial use | Allowed | Allowed |
| Languages | — | 1 |
| Downloads 30d | 5,227,079 | 805,145 |
| Downloads all time | 75.8 M | 20.5 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2025-10-11 | 2024-08-16 |
| API input price / 1M tokens | $0.25 | not on OpenRouter |
| API output price / 1M tokens | $0.75 | not on OpenRouter |
| Est. cost per 1M input + 1M output tokens | $1.00 | not on OpenRouter |
| Cached input / 1M tokens | $0.12 | not on OpenRouter |
| AA Intelligence Index | 7.3 | not on OpenRouter |
| AA Coding Index | — | not on OpenRouter |
| AA Agentic Index | — | not on OpenRouter |
| Max output tokens | 235,929 | not on OpenRouter |
| API providers | 1 | not on OpenRouter |
| Tool calling | Yes | not on OpenRouter |
| Reasoning mode | No | not on OpenRouter |
API cost
Estimated cost of 1M input + 1M output tokens via OpenRouter
Qwen3-VL-8B-Instruct
$1.00
Download trend
Daily snapshots, last 90 days (28 Jul – 25 Oct)
20.2 M796 K
Qwen3-VL-8B-Instruct
Phi-3.5-vision-instruct
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
Phi-3.5-vision-instruct · fp16 @ 128K ctx
19.6 GB / 24 GB
Qwen3-VL-8B-Instruct · fp16
21.1 GB / 24 GB