RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Qwen: Qwen3.5-9B API pricing

$0.17 per 1M input tokens and $0.25 per 1M output tokens on OpenRouter. Context window: 262,144 tokens.

$0.17input / 1M tokens
$0.25output / 1M tokens
$0.153cached input / 1M
256Kcontext
text+image+video->textmodality

What it costs in practice

WorkloadCost
Chatbot: 10,000 messages (1,500 in / 400 out tokens each)$3.55
RAG: 1,000 queries (8,000 in / 600 out each)$1.51
Summarize 100 long documents (50,000 in / 1,000 out each)$0.875

Cross-check against the provider's list price

No first-party entry for this model in LiteLLM's price map, so this price comes from OpenRouter only.

Price history

When (UTC)FieldOldNew
2026-11-17 04:17input$0.099$0.068
2026-11-17 04:17output$0.146$0.10
2026-11-17 04:17cacheRead$0.089$0.061
2026-11-17 04:19input$0.068$0.073
2026-11-17 04:19output$0.10$0.107
2026-11-17 04:19cacheRead$0.061$0.065
2026-11-17 06:13input$0.073$0.17
2026-11-17 06:13output$0.107$0.25
2026-11-17 06:13cacheRead$0.065$0.153
2026-11-17 12:43input$0.17$0.149
2026-11-17 12:43output$0.25$0.22
2026-11-17 12:43cacheRead$0.153$0.135
2026-11-17 13:28input$0.149$0.17
2026-11-17 13:28output$0.22$0.25
2026-11-17 13:28cacheRead$0.135$0.153

Run it yourself instead

The open weights are on Hugging Face. See the VRAM Qwen3.5 9B needs and which GPUs fit it →

Cheaper alternatives with similar context

Text models with at least 125K context and a lower blended price (75% input, 25% output). Price only, not a quality ranking.

ModelInput / 1MOutput / 1MContext
PrismML: Ternary Bonsai 2 27B$0.075$0.50256K
Meta: Llama Guard 4 12B$0.18$0.18160K
Xiaomi: MiMo-V2.6-Flash$0.14$0.281.0013580322265625M
ByteDance Seed: Seed-2.0-Mini$0.10$0.40256K
Google: Gemini 2.5 Flash Lite$0.10$0.401M
OpenAI: GPT-4.1 Nano$0.10$0.401023K
Particle AI: Particle-1$0.06$0.501M
DeepSeek: DeepSeek V4.1 Flash (batch)$0.112$0.3361M

Source: OpenRouter listing via the public models API. Prices checked Nov 17, 2026. Model ID qwen/qwen3.5-9b, listed Mar 10, 2026. Report an error.