Qwen2.5 VL 72B Instruct vs Qwen3 VL 8B Instruct
Qwen3 VL 8B Instruct costs 1.3x less per output token than Qwen2.5 VL 72B Instruct ($0.75 vs $1.00 per million). Updated October 22, 2026; no benchmark has tested both models yet.
- Coding
- No data
- Overall
- No data
- Output $/M
- $1.00
- Coding
- No data
- Overall
- No data
- Output $/M
- $0.75
Qwen2.5 VL 72B Instruct or Qwen3 VL 8B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Compare them on price, context and the boards below.
No benchmark has tested both yet. Qwen2.5 VL 72B Instruct has results on OSWorld, Kagi LLM Benchmark and GeoBench; Qwen3 VL 8B Instruct on LiveCodeBench (Artificial Analysis run), Terminal-Bench Hard (Artificial Analysis run) and AA Intelligence Index. Compare them on price and context instead.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
No call
Only Qwen3 VL 8B Instruct has a AA Intelligence Index result, 7.3.
Directional only
Long documents
Listed context window
Qwen3 VL 8B Instruct
Qwen3 VL 8B Instruct takes 262K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3 VL 8B Instruct
Qwen2.5 VL 72B Instruct $0.0013, Qwen3 VL 8B Instruct $0.0006 per turn.
List prices
At a glance
- Coding index
- Qwen2.5 VL 72B Instruct not benchmarked, Qwen3 VL 8B Instruct not benchmarked.
- Widest category
- None yet: no category has been tested on both models.
- Widest single gap
- None yet: no benchmark has tested both models.
- Output price
- Qwen2.5 VL 72B Instruct $1.00 against Qwen3 VL 8B Instruct $0.75 per million output tokens, 1.3x apart.
- Context window
- Qwen2.5 VL 72B Instruct takes 128K tokens, Qwen3 VL 8B Instruct 262K.
Benchmark by benchmark
Every board either model has a result on, a category to a fold; bars are each result on a 0 to 100 scale. Only one model is measured on each, so no row names a leader.
CodingDirectional onlySix of these boards feed the Vibe Coding Index2 boards measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| LiveCodeBench (Artificial Analysis run) | Not measured | 35.3% | Only Qwen3 VL 8B Instruct measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 3.8% | Only Qwen3 VL 8B Instruct measured |
| Neither model has a result on the other 41 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 7.3 | Only Qwen3 VL 8B Instruct measured |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| τ²-Bench Telecom | Not measured | 29.2% | Only Qwen3 VL 8B Instruct measured |
| OSWorld | 5.0% | Not measured | Only Qwen2.5 VL 72B Instruct measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index3 boards measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 57.9% | Only Qwen3 VL 8B Instruct measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 3.8% | Only Qwen3 VL 8B Instruct measured |
| Kagi LLM Benchmark | 36.0% | Not measured | Only Qwen2.5 VL 72B Instruct measured |
| Neither model has a result on the other 20 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index2 boards measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| AA Math Index | Not measured | 30.7 | Only Qwen3 VL 8B Instruct measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 30.7% | Only Qwen3 VL 8B Instruct measured |
| Neither model has a result on the other 12 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index2 boards measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | Not measured | 74.9% | Only Qwen3 VL 8B Instruct measured |
| GeoBench | 62.0% | Not measured | Only Qwen2.5 VL 72B Instruct measured |
| Neither model has a result on the other 18 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index5 boards measured
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct | Leader |
|---|---|---|---|
| IFBench | Not measured | 39.9% | Only Qwen3 VL 8B Instruct measured |
| Long context reasoning (AA-LCR) | Not measured | 33.3% | Only Qwen3 VL 8B Instruct measured |
| LMArena Vision | 1107 | Not measured | Only Qwen2.5 VL 72B Instruct measured |
| LMArena Vision: English | 1115 | Not measured | Only Qwen2.5 VL 72B Instruct measured |
| LMArena Vision: Chinese | 1107 | Not measured | Only Qwen2.5 VL 72B Instruct measured |
| Neither model has a result on the other 43 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Qwen3 VL 8B Instruct costs less
Repository review
50K input + 3K output tokens
Qwen3 VL 8B Instruct costs less
Agent loop
200K input + 10K output tokens
Qwen3 VL 8B Instruct costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Qwen2.5 VL 72B Instruct | Qwen3 VL 8B Instruct |
|---|---|---|
| Maker | Qwen | Qwen |
| Weights | Open weights | Open weights |
| Released | February 1, 2025 | October 14, 2025 |
| Context window | 128K | 262K |
| Input $/Mtok | $0.80 | $0.25 |
| Output $/Mtok | $1.00 | $0.75 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen2.5 VL 72B Instruct better than Qwen3 VL 8B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Compare them on price, context and the boards below.
Which is cheaper, Qwen2.5 VL 72B Instruct or Qwen3 VL 8B Instruct?
Qwen2.5 VL 72B Instruct costs $0.80 per million input tokens and $1.00 per million output tokens; Qwen3 VL 8B Instruct costs $0.25 and $0.75. Cost per solved task is known for neither of them.
Which has the bigger context window, Qwen2.5 VL 72B Instruct or Qwen3 VL 8B Instruct?
Qwen3 VL 8B Instruct, with 262K tokens against 128K.
Has any benchmark tested both Qwen2.5 VL 72B Instruct and Qwen3 VL 8B Instruct?
No benchmark has tested both yet. Qwen2.5 VL 72B Instruct has results on OSWorld, Kagi LLM Benchmark and GeoBench; Qwen3 VL 8B Instruct on LiveCodeBench (Artificial Analysis run), Terminal-Bench Hard (Artificial Analysis run) and AA Intelligence Index. Compare them on price and context instead. Data as of October 22, 2026.