Model comparison
Mistral Small 3.1 vs Qwen2.5-VL
- Input
- $0.351 / 1M tokens
- Output
- $0.555 / 1M tokens
- Context
- 128K
- Released
- 2025-03-17
- Input
- $0.80 per 1M input tokens
- Output
- $1.00 per 1M output tokens
- Context
- 32,768 tokens (extendable to ~64K for long video via YaRN; some API providers expose up to 131K)
- Released
- 2025-01-26
Qwen2.5-VL scores higher on 4 of the 4 benchmarks both models report; Mistral Small 3.1 is cheaper on input tokens ($0.4 vs $0.8 per 1M, 2.3×); Mistral Small 3.1 has the larger context window (128K vs 33K).
Specs and pricing side by side
| Spec | Mistral Small 3.1 | Qwen2.5-VL |
|---|---|---|
| Maker | Mistral AI | Alibaba (Qwen) |
| Released | 2025-03-17 | 2025-01-26 |
| Context window | 128K | 32,768 tokens (extendable to ~64K for long video via YaRN; some API providers expose up to 131K) |
| Max output | not published | Not separately documented; bounded by the… |
| Input price | $0.351 / 1M tokens | $0.80 per 1M input tokens |
| Output price | $0.555 / 1M tokens | $1.00 per 1M output tokens |
| Parameters | 24B | Four open sizes: 3B, 7B, 32B and 72B (each in base and instruction-tuned variants) |
| License | Apache 2.0 | Mixed by size: Qwen2.5-VL-72B-Instruct under the Qwen License; 7B and 32B under Apache 2.0; 3B under the Qwen Research License (non-commercial only). |
| Weights | Open weights | Open weightssame |
| Modalities | Text, Vision | text, image, video |
| Knowledge cutoff | October 2023 | Not officially published by Alibaba |
Benchmarks both models report
| Benchmark | Mistral Small 3.1 | Qwen2.5-VL |
|---|---|---|
| MMMU | 64% | 70.2% |
| MathVista | 68.91% | 74.8% |
| ChartQA | 86.24% | 89.5% |
| DocVQA | 94.08% | 96.4% |
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 Mistral Small 3.1 if…
- Cost matters: input tokens cost 2.3× less ($0.4 vs $0.8 per 1M).
- You need the longer context (128K vs 33K).
- You need Text / Vision input, which Qwen2.5-VL lacks.
- Your use case is: self-hosted multimodal assistants that need to read images, documents, and charts.
Pick Qwen2.5-VL if…
- You care about MMMU, MathVista, ChartQA (it scores higher on 4 shared benchmarks).
- You need text / image / video input, which Mistral Small 3.1 lacks.
- Your use case is: document AI: extracting text, tables and layout from invoices, contracts, research papers and scanned forms.
FAQ
Which is better, Mistral Small 3.1 or Qwen2.5-VL?
Qwen2.5-VL scores higher on 4 of the 4 benchmarks both models report; Mistral Small 3.1 is cheaper on input tokens ($0.4 vs $0.8 per 1M, 2.3×); Mistral Small 3.1 has the larger context window (128K vs 33K). The better pick depends on whether you value benchmark scores, price per token, context length or openness for your workload.
How much cheaper is Mistral Small 3.1?
Mistral Small 3.1: $0.351 / 1M tokens input, $0.555 / 1M tokens output. Qwen2.5-VL: $0.80 per 1M input tokens input, $1.00 per 1M output tokens output (list API prices as published by each maker).
What are the context windows of Mistral Small 3.1 and Qwen2.5-VL?
Mistral Small 3.1 offers 128K; Qwen2.5-VL offers 32,768 tokens (extendable to ~64K for long video via YaRN; some API providers expose up to 131K) with up to Not separately documented; bounded by the 32,768-token context window output.
Is Mistral Small 3.1 newer than Qwen2.5-VL?
Mistral Small 3.1 was released 2025-03-17; Qwen2.5-VL was released 2025-01-26.
Go deeper
- Mistral Small 3.1: full specs, pricing and benchmarks
- Qwen2.5-VL: full specs, pricing and benchmarks
- mistral small 3 2 vs mistral small 3 1
- qwen3 vl vs qwen2 5 vl
- mistral small 3 1 vs mistral small 3
- kimi vl a3b thinking 2506 vs qwen2 5 vl
- mistral small 3 1 vs llama 3 3
- qwen2 5 vl vs pixtral 12b
- All model comparisons