Alternatives · 15 models compared · checked 10 Oct 2026
Qwen2.5 VL 72B Instruct alternatives: 15 models in or near grid D2
The closest model to Qwen2.5 VL 72B Instruct on the atlas is MiniMax M2-her, in the same square, at $0.30 per 1M input tokens against Qwen2.5 VL 72B Instruct’s . 15 of the 15 nearest models are cheaper on input; the cheapest is Qwen2.5 7B Instruct at $0.10. We rank by position on the atlas, not by capability, because we hold no capability data.
$0.80
observed We measured
- Source
- Our catalogue
- Retrieved
- 10 Oct 2026
Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
Cite
Cite this page
ModelsAtlas. “Qwen2.5 VL 72B Instruct alternatives: 15 models in or near grid D2”. Retrieved 10 Oct 2026. https://modelsatlas.com/models/qwen/2-5-vl-72b-instruct/alternatives
Where the alternatives sit
Each square is a price band (A–H) × a context band (1–6). How the grid works
Same squareGRIDD26 modelsMiniMax M2-her, Mistral Small 3.1 24B, Nano Banana (Gemini 2.5 Flash Image) and 3 more
Qwen2.5 VL 72B InstructIts 15 nearest modelsOther modelsOpen weights
View the map as a table (3 squares)
| Square | Price band | Context band | Models |
|---|---|---|---|
| GRIDC2 | $0.10–0.30 | 32–128K | Qwen2.5 7B Instruct, Voxtral Small 24B 2507, Qwen3 14B, Qwen3 30B A3B, Command R (08-2024), Saba, Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) |
| GRIDD2 | $0.30–1 | 32–128K | Qwen2.5 VL 72B Instruct, MiniMax M2-her, Nano Banana (Gemini 2.5 Flash Image), Mistral Small 3.1 24B, Qwen2.5 72B Instruct, WizardLM-2 8x22B, Qwen2.5 Coder 32B Instruct |
| GRIDD3 | $0.30–1 | 128–256K | Codestral 2508, Command A+ |
Top 15 alternatives
Ranked by steps away on the atlas, then cheaper before pricier, then input price
| # | Model | Grid | Why it fits | Input / 1M | vs Qwen2.5 VL 72B Instruct | Output / 1M | Context | |
|---|---|---|---|---|---|---|---|---|
| — | Qwen2.5 VL 72B InstructQwen (Alibaba) · the model you’re replacing | GRIDD2 | Baseline | $0.80 | — | $1.00 | 128K | |
| 1 | MiniMax M2-herMiniMax | GRIDD2 | Same square, 63% cheaper input | $0.30 | −62% | $1.20 | 66K | Compare → |
| 2 | Nano Banana (Gemini 2.5 Flash Image)Google | GRIDD2 | Same square, 63% cheaper input | $0.30 | −62% | $2.50 | 33K | Compare → |
| 3 | Mistral Small 3.1 24BMistral AI | GRIDD2 | Same square, 56% cheaper input | $0.35 | −56% | $0.56 | 128K | Compare → |
| 4 | Qwen2.5 72B InstructQwen (Alibaba) | GRIDD2 | Same maker, same square, 55% cheaper input | $0.36 | −55% | $0.40 | 33K | Compare → |
| 5 | WizardLM-2 8x22BMicrosoft | GRIDD2 | Same square, 23% cheaper input | $0.62 | −23% | $0.62 | 66K | Compare → |
| 6 | Qwen2.5 Coder 32B InstructQwen (Alibaba) | GRIDD2 | Same maker, same square, 18% cheaper input | $0.66 | −17% | $1.00 | 33K | Compare → |
| 7 | Qwen2.5 7B InstructQwen (Alibaba) · open weights | GRIDC2 | Same maker, neighbouring square C2, 88% cheaper input, open weights | $0.10 | −87% | $0.20 | 33K | |
| 8 | Voxtral Small 24B 2507Mistral AI | GRIDC2 | Neighbouring square C2, 88% cheaper input | $0.10 | −87% | $0.30 | 33K | |
| 9 | Qwen3 14BQwen (Alibaba) · open weights | GRIDC2 | Same maker, neighbouring square C2, 85% cheaper input, open weights | $0.12 | −85% | $0.24 | 41K | |
| 10 | Qwen3 30B A3BQwen (Alibaba) | GRIDC2 | Same maker, neighbouring square C2, 85% cheaper input | $0.12 | −85% | $0.50 | 41K | |
| 11 | Command R (08-2024)Cohere | GRIDC2 | Neighbouring square C2, 81% cheaper input | $0.15 | −81% | $0.60 | 128K | |
| 12 | SabaMistral AI | GRIDC2 | Neighbouring square C2, 75% cheaper input | $0.20 | −75% | $0.60 | 33K | |
| 13 | Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)Google | GRIDC2 | Neighbouring square C2, 69% cheaper input | $0.25 | −69% | $1.50 | 66K | |
| 14 | Codestral 2508Mistral AI | GRIDD3 | Neighbouring square D3, 63% cheaper input, 2× the context | $0.30 | −62% | $0.90 | 256K | |
| 15 | Command A+Cohere | GRIDD3 | Neighbouring square D3, 63% cheaper input, 1.5× the context | $0.30 | −62% | $1.50 | 192K |
What switching saves
3,000 input + 800 output tokens × 1,000 requests a day
Σ Derived from standard-tier prices; excludes caching and batch discounts. Use your own numbers →
Larger-context alternatives
| Model | Input / 1M | Context | |
|---|---|---|---|
| Codestral 2508Mistral AI | $0.30 | 256K | |
| Command A+Cohere | $0.30 | 192K | |
| MiniMax M2MiniMax | $0.30 | 205K | |
| MiniMax M2.1MiniMax | $0.30 | 205K | |
| Muse Glimmer 30BMeta | $0.30 | 131K |
Free alternatives
| Model | Input / 1M | Context | |
|---|---|---|---|
| Lyria 3 Clip PreviewGoogle | Free | 1.05M | |
| Lyria 3 Pro PreviewGoogle | Free | 1.05M |
Compare Qwen2.5 VL 72B Instruct head to head
Questions
What is the best alternative to Qwen2.5 VL 72B Instruct?
MiniMax M2-her ranks first: same square, 63% cheaper input. Input is $0.30 per 1M tokens against $0.80 for Qwen2.5 VL 72B Instruct. Ranking is by position on the atlas, not by quality.
What is the cheapest alternative to Qwen2.5 VL 72B Instruct?
Qwen2.5 7B Instruct at $0.10 per 1M input tokens, 88% less than Qwen2.5 VL 72B Instruct, among the 15 closest.
Is there an open-weight alternative to Qwen2.5 VL 72B Instruct?
Yes, 2 of the 15: Qwen2.5 7B Instruct and Qwen3 14B.
Which alternative has a larger context window?
Codestral 2508 and Command A+ hold more than Qwen2.5 VL 72B Instruct's 128K; the largest is Codestral 2508 at 256K.
How are alternatives ranked?
By distance on the atlas grid from Qwen2.5 VL 72B Instruct's square D2: the same square first, then neighbouring squares. Within the same distance, cheaper models come before pricier ones, then the lowest input price wins. We hold no capability data, so we do not claim a model can do what Qwen2.5 VL 72B Instruct does, only that it costs and holds about the same. No vendor pays for position.
More alternatives
Keep going in your AI
Opens with this page as the source. We add nothing else to the prompt.
- What is the best alternative to Qwen2.5 VL 72B Instruct?
- What is the cheapest alternative to Qwen2.5 VL 72B Instruct?
- Is there an open-weight alternative to Qwen2.5 VL 72B Instruct?