Claude Haiku 4.5 vs Qwen3 VL 8B Thinking
Side-by-side API pricing and Artificial Analysis performance for Claude Haiku 4.5 (Anthropic) and Qwen3 VL 8B Thinking (Alibaba (Qwen)). Green cells mark the better value in each row.
Prices updated Jul 22, 2026, 1:35 AM UTC · refreshed hourly
| Claude Haiku 4.5 Anthropic | Qwen3 VL 8B Thinking Alibaba (Qwen) | |
|---|---|---|
| Pricing | ||
| Input $/1M | $1.00 | $0.117 |
| Output $/1M | $5.00 | $1.36 |
| Blended $/1M (3:1) | $2.00 | $0.429 |
| RAG example (30K in / 2K out) | $0.04 | $0.00624 |
| Quality & speed | ||
| Intelligence Index | 29.6 | — |
| Coding Index | 43.9 | — |
| Agentic Index | 16.4 | — |
| Output speed (tok/s) | — | — |
| Time to first token | — | — |
| Specs | ||
| Context window | 200K | 131K |
| Input modalities | text, image, file | image, text |
FAQ
Which is cheaper, Claude Haiku 4.5 or Qwen3 VL 8B Thinking?
Qwen3 VL 8B Thinking has the lower blended API price at $0.429 per 1M tokens (3:1 input:output mix), versus $2.00 for the other model. Prices are live from OpenRouter and refresh about hourly.
Which is better for RAG workloads?
For a typical RAG request (30K input / 2K output tokens), Qwen3 VL 8B Thinking costs about $0.00624 per request versus $0.04. Also compare context windows: Claude Haiku 4.5 offers 200K and Qwen3 VL 8B Thinking offers 131K.
Which scores higher on Artificial Analysis benchmarks?
Claude Haiku 4.5 leads on the Artificial Analysis Intelligence Index (29.6 vs —). Check coding and agentic indices on this page for workload-specific tradeoffs.