{"benchmark_id":"charxiv-d","name":"CharXiv-D","parent_benchmark":null,"categories":["multimodal","reasoning","structured_output","vision"],"modality":"multimodal","multilingual":false,"max_score":1.0,"language":"en","description":"CharXiv-D is the descriptive questions subset of the CharXiv benchmark, designed to assess multimodal large language models' ability to extract basic information from scientific charts. It contains descriptive questions covering information extraction, enumeration, pattern recognition, and counting across 2,323 diverse charts from arXiv papers, all curated and verified by human experts.","paper_link":"https://arxiv.org/abs/2406.18521","implementation_link":null,"verified":false,"created_at":"2026-05-07T16:53:22.318796+00:00","updated_at":"2026-07-05T18:27:45.140792+00:00","statistics":{"total_models":16,"average_score":0.8660625000000001,"min_score":0.739,"max_score":0.955,"score_stddev":0.05674559454266028,"verified_count":0,"self_reported_count":16},"child_benchmarks":[],"linked_dataset":null,"models":[{"rank":1,"model_id":"seed-2.1-pro","model_name":"Seed 2.1 Pro","organization_id":"bytedance","organization_name":"ByteDance","organization_country":"CN","score":0.955,"normalized_score":0.955,"verified":false,"self_reported":true,"self_reported_source":"https://seed.bytedance.com/en/seed2_1","analysis_method":null,"verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2026-06-24","param_count":null,"is_open_source":false,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":2,"model_id":"seed-2.1-turbo","model_name":"Seed 2.1 Turbo","organization_id":"bytedance","organization_name":"ByteDance","organization_country":"CN","score":0.946,"normalized_score":0.946,"verified":false,"self_reported":true,"self_reported_source":"https://seed.bytedance.com/en/seed2_1","analysis_method":null,"verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2026-06-24","param_count":null,"is_open_source":false,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":3,"model_id":"qwen3-vl-32b-instruct","model_name":"Qwen3 VL 32B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.905,"normalized_score":0.905,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DOI","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-09-22","param_count":33000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":4,"model_id":"qwen3-vl-32b-thinking","model_name":"Qwen3 VL 32B Thinking","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.902,"normalized_score":0.902,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":null,"verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-09-22","param_count":33000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":5,"model_id":"gpt-4.5","model_name":"GPT-4.5","organization_id":"openai","organization_name":"OpenAI","organization_country":"US","score":0.9,"normalized_score":0.9,"verified":false,"self_reported":true,"self_reported_source":"https://openai.com/index/gpt-4-1/","analysis_method":"Accuracy","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-02-27","param_count":null,"is_open_source":false,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":6,"model_id":"gpt-4.1-mini-2025-04-14","model_name":"GPT-4.1 mini","organization_id":"openai","organization_name":"OpenAI","organization_country":"US","score":0.884,"normalized_score":0.884,"verified":false,"self_reported":true,"self_reported_source":"https://openai.com/index/introducing-gpt-5-for-developers/","analysis_method":"Standard benchmark","verification_date":null,"provider_id":"openai","input_cost_per_million":0.4,"output_cost_per_million":1.6,"context_window":1047576,"announcement_date":"2025-04-14","param_count":null,"is_open_source":false,"is_new":false,"best_latency":5.0,"latency_provider":"OpenAI","best_throughput":150.0,"throughput_provider":"OpenAI","context_provider":"OpenAI"},{"rank":7,"model_id":"command-a-plus-05-2026","model_name":"Command A+","organization_id":"cohere","organization_name":"Cohere","organization_country":"CA","score":0.88,"normalized_score":0.88,"verified":false,"self_reported":true,"self_reported_source":"https://cohere.com/blog/command-a-plus","analysis_method":"CharXiv descriptive; standard methodology reported by Cohere","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2026-05-20","param_count":218000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":8,"model_id":"gpt-4.1-2025-04-14","model_name":"GPT-4.1","organization_id":"openai","organization_name":"OpenAI","organization_country":"US","score":0.879,"normalized_score":0.879,"verified":false,"self_reported":true,"self_reported_source":"https://openai.com/index/introducing-gpt-5-for-developers/","analysis_method":"Standard benchmark","verification_date":null,"provider_id":"openai","input_cost_per_million":2.0,"output_cost_per_million":8.0,"context_window":1047576,"announcement_date":"2025-04-14","param_count":null,"is_open_source":false,"is_new":false,"best_latency":10.0,"latency_provider":"OpenAI","best_throughput":100.0,"throughput_provider":"OpenAI","context_provider":"OpenAI"},{"rank":9,"model_id":"qwen3-vl-30b-a3b-thinking","model_name":"Qwen3 VL 30B A3B Thinking","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.869,"normalized_score":0.869,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DQ","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-09-22","param_count":31000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":10,"model_id":"qwen3-vl-8b-thinking","model_name":"Qwen3 VL 8B Thinking","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.859,"normalized_score":0.859,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DQ","verification_date":null,"provider_id":"deepinfra","input_cost_per_million":0.18,"output_cost_per_million":2.09,"context_window":262144,"announcement_date":"2025-09-22","param_count":9000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":"DeepInfra","best_throughput":null,"throughput_provider":"DeepInfra","context_provider":"DeepInfra"},{"rank":11,"model_id":"qwen3-vl-30b-a3b-instruct","model_name":"Qwen3 VL 30B A3B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.855,"normalized_score":0.855,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DQ","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-09-22","param_count":31000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":12,"model_id":"gpt-4o-2024-08-06","model_name":"GPT-4o","organization_id":"openai","organization_name":"OpenAI","organization_country":"US","score":0.853,"normalized_score":0.853,"verified":false,"self_reported":true,"self_reported_source":"https://openai.com/index/gpt-4-1/","analysis_method":"Accuracy","verification_date":null,"provider_id":"openai","input_cost_per_million":2.5,"output_cost_per_million":10.0,"context_window":128000,"announcement_date":"2024-08-06","param_count":null,"is_open_source":false,"is_new":false,"best_latency":0.5,"latency_provider":"OpenAI","best_throughput":132.0,"throughput_provider":"OpenAI","context_provider":"OpenAI"},{"rank":13,"model_id":"qwen3-vl-4b-thinking","model_name":"Qwen3 VL 4B Thinking","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.839,"normalized_score":0.839,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DQ","verification_date":null,"provider_id":"deepinfra","input_cost_per_million":0.1,"output_cost_per_million":1.0,"context_window":262144,"announcement_date":"2025-09-22","param_count":4000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":"DeepInfra","best_throughput":null,"throughput_provider":"DeepInfra","context_provider":"DeepInfra"},{"rank":14,"model_id":"qwen3-vl-8b-instruct","model_name":"Qwen3 VL 8B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.83,"normalized_score":0.83,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DQ","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2025-09-22","param_count":9000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":15,"model_id":"qwen3-vl-4b-instruct","model_name":"Qwen3 VL 4B Instruct","organization_id":"qwen","organization_name":"Alibaba Cloud / Qwen Team","organization_country":"CN","score":0.762,"normalized_score":0.762,"verified":false,"self_reported":true,"self_reported_source":null,"analysis_method":"DQ","verification_date":null,"provider_id":"deepinfra","input_cost_per_million":0.1,"output_cost_per_million":0.6,"context_window":262144,"announcement_date":"2025-09-22","param_count":4000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":"DeepInfra","best_throughput":null,"throughput_provider":"DeepInfra","context_provider":"DeepInfra"},{"rank":16,"model_id":"gpt-4.1-nano-2025-04-14","model_name":"GPT-4.1 nano","organization_id":"openai","organization_name":"OpenAI","organization_country":"US","score":0.739,"normalized_score":0.739,"verified":false,"self_reported":true,"self_reported_source":"https://openai.com/index/gpt-4-1/","analysis_method":"Standard benchmark","verification_date":null,"provider_id":"openai","input_cost_per_million":0.1,"output_cost_per_million":0.4,"context_window":1047576,"announcement_date":"2025-04-14","param_count":null,"is_open_source":false,"is_new":false,"best_latency":2.0,"latency_provider":"OpenAI","best_throughput":200.0,"throughput_provider":"OpenAI","context_provider":"OpenAI"}]}