BenchmarksAlibabaToday
Qwen3 VL 235B A22B Instruct Benchmark Update
Quality: 14.3/100 | Price: $1.225/M tokens | Output: 49.561 tok/s | MMLU: 0.823% | HumanEval: 0.594%
View sourceQwen
Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table...
Running this yourself: likely needs a high-memory cloud gpu.
41.9
Quality Score
1215
Arena ELO
235B
Parameters
262K
Context
Sign in to join the discussion
0
Downloads
0
Likes
Sep 2025
Released
Benchmarks
19
Open Source
1
Recent launch, pricing, benchmark, and API signals linked to this model or its provider.
Quality: 14.3/100 | Price: $1.225/M tokens | Output: 49.561 tok/s | MMLU: 0.823% | HumanEval: 0.594%
View sourceQuality: 14.3/100 | Price: $1.225/M tokens | Output: 49.561 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $1.225/M tokens | Output: 50.482 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 50.406 tok/s | MMLU: 0.823% | HumanEval: 0.594%
View sourceQuality: 14.3/100 | Price: $0.7/M tokens | Output: 50.721 tok/s | MMLU: 0.823% | HumanEval: 0.594%
View sourceQuality: 14.3/100 | Price: $0.7/M tokens | Output: 50.302 tok/s | MMLU: 0.823% | HumanEval: 0.594%
View sourceQuality: 14.3/100 | Price: $1.225/M tokens | Output: 50.482 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 50.406 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 50.721 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 50.302 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 50.562 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 49.811 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 51.391 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 52.456 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 51.769 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 51.957 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 52.27 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 54.288 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 55.467 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 54.689 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 43.883 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 47.824 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 48.173 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Quality: 14.3/100 | Price: $0.7/M tokens | Output: 50.731 tok/s | MMLU: 0.823% | HumanEval: 0.594%
Qwen3 VL 235B A22B Instruct is now available through local Ollama runtime and Ollama Cloud. 256K context window listed. The most powerful vision-language model in the Qwen model family to date.