Model comparison
Qwen3-VL vs Kimi-VL-A3B-Thinking-2506
- Input
- $0.20 / 1M tokens (0-32K tier)
- Output
- $1.60 / 1M tokens (0-32K tier)
- Context
- 256K
- Released
- 2025-09-23
- Input
- $0.025 / 1M tokens
- Output
- $0.10 / 1M tokens
- Context
- 128K
- Released
- 2025-06-21
Qwen3-VL scores higher on 4 of the 4 benchmarks both models report; Kimi-VL-A3B-Thinking-2506 is cheaper on input tokens ($0 vs $0.2 per 1M, 8×); Qwen3-VL has the larger context window (256K vs 128K).
Specs and pricing side by side
| Spec | Qwen3-VL | Kimi-VL-A3B-Thinking-2506 |
|---|---|---|
| Maker | Alibaba (Qwen) | Moonshot AI (Kimi) |
| Released | 2025-09-23 | 2025-06-21 |
| Context window | 256K | 128K |
| Max output | 32K | 32Ksame |
| Input price | $0.20 / 1M tokens (0-32K tier) | $0.025 / 1M tokens |
| Output price | $1.60 / 1M tokens (0-32K tier) | $0.10 / 1M tokens |
| Parameters | 235B total / 22B active (flagship MoE); also dense 2B/4B/8B/32B and MoE 30B-A3B | 16B total / 2.8B active (MoE) |
| License | Apache 2.0 | MIT |
| Weights | Open weights | Open weightssame |
| Modalities | Text, Vision, Video, PDF | Text, Vision, Video, PDFsame |
| Knowledge cutoff | Not publicly disclosed | December 2024 |
Benchmarks both models report
| Benchmark | Qwen3-VL | Kimi-VL-A3B-Thinking-2506 |
|---|---|---|
| MathVista | 85.8% | 80.1% |
| MathVision | 74.6% | 56.9% |
| MMMU-Pro | 69.3% | 46.3% |
| ScreenSpot Pro | 61.8% | 52.8% |
Scores are as published by each maker or its cited source; harness, reasoning effort and date can differ between rows, so treat small gaps as ties. Sources are listed on each model page.
Which one should you pick?
Pick Qwen3-VL if…
- You care about MathVista, MathVision, MMMU-Pro (it scores higher on 4 shared benchmarks).
- You need the longer context (256K vs 128K).
- Your use case is: document, invoice, and form parsing with multilingual OCR.
Pick Kimi-VL-A3B-Thinking-2506 if…
- Cost matters: input tokens cost 8× less ($0 vs $0.2 per 1M).
- Your use case is: visual and multimodal math/reasoning over charts, diagrams and screenshots.
FAQ
Which is better, Qwen3-VL or Kimi-VL-A3B-Thinking-2506?
Qwen3-VL scores higher on 4 of the 4 benchmarks both models report; Kimi-VL-A3B-Thinking-2506 is cheaper on input tokens ($0 vs $0.2 per 1M, 8×); Qwen3-VL has the larger context window (256K vs 128K). The better pick depends on whether you value benchmark scores, price per token, context length or openness for your workload.
How much cheaper is Kimi-VL-A3B-Thinking-2506?
Qwen3-VL: $0.20 / 1M tokens (0-32K tier) input, $1.60 / 1M tokens (0-32K tier) output. Kimi-VL-A3B-Thinking-2506: $0.025 / 1M tokens input, $0.10 / 1M tokens output (list API prices as published by each maker).
What are the context windows of Qwen3-VL and Kimi-VL-A3B-Thinking-2506?
Qwen3-VL offers 256K with up to 32K output; Kimi-VL-A3B-Thinking-2506 offers 128K with up to 32K output.
Is Qwen3-VL newer than Kimi-VL-A3B-Thinking-2506?
Qwen3-VL was released 2025-09-23; Kimi-VL-A3B-Thinking-2506 was released 2025-06-21.