{"run_id":"run-20260715T040307-80bf67","started_at":1784088187,"summary":{"run_id":"run-20260715T040307-80bf67","question_bank_version":"2026-05-17-v1","code_version":"0.2.6","started_at":1784088187.434489,"judges_active":["gpt-5","claude-opus-4-8","claude-opus-4-7","mistral-large-2"],"by_model_domain":[{"model_id":"claude-opus-4-7","domain":"jura","score_mean":0.8085572562358276,"score_ci_lo":0.7599007936507937,"score_ci_hi":0.8547675736961451,"top100":81,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.8085572562358276,"hallu_rate_mean":0.9071428571428571,"hallu_prevalence":0.8285714285714286,"hallu_top100":77,"inter_rater_agreement":0.7655668934240363,"n_judges":4,"tokens_in_total":22950,"tokens_out_total":31109,"model_cost_usd":0.892475},{"model_id":"claude-opus-4-7","domain":"medizin","score_mean":0.8426587301587303,"score_ci_lo":0.7650545634920636,"score_ci_hi":0.9074156746031747,"top100":84,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.8426587301587303,"hallu_rate_mean":1.15,"hallu_prevalence":1,"hallu_top100":71,"inter_rater_agreement":0.8643849206349206,"n_judges":4,"tokens_in_total":11495,"tokens_out_total":20534,"model_cost_usd":0.570825},{"model_id":"claude-opus-4-7","domain":"steuerrecht","score_mean":0.7216005291005291,"score_ci_lo":0.6479431216931216,"score_ci_hi":0.794074074074074,"top100":72,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.7216005291005291,"hallu_rate_mean":2.65,"hallu_prevalence":1,"hallu_top100":33,"inter_rater_agreement":0.7889153439153439,"n_judges":4,"tokens_in_total":9810,"tokens_out_total":19251,"model_cost_usd":0.530325},{"model_id":"claude-opus-4-7","domain":"wirtschaftsrecht","score_mean":0.7803009259259259,"score_ci_lo":0.7373445767195768,"score_ci_hi":0.8297321428571428,"top100":78,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.7803009259259259,"hallu_rate_mean":2.2666666666666666,"hallu_prevalence":0.9666666666666667,"hallu_top100":43,"inter_rater_agreement":0.7409920634920635,"n_judges":4,"tokens_in_total":22262,"tokens_out_total":41881,"model_cost_usd":1.158335},{"model_id":"claude-opus-4-8","domain":"jura","score_mean":0.8077777777777778,"score_ci_lo":0.7674007936507937,"score_ci_hi":0.8472619047619047,"top100":81,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.8077777777777778,"hallu_rate_mean":0.6642857142857143,"hallu_prevalence":0.8,"hallu_top100":83,"inter_rater_agreement":0.816201814058957,"n_judges":4,"tokens_in_total":22775,"tokens_out_total":32705,"model_cost_usd":0.9315},{"model_id":"claude-opus-4-8","domain":"medizin","score_mean":0.801264880952381,"score_ci_lo":0.7326140873015873,"score_ci_hi":0.8603174603174603,"top100":80,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.801264880952381,"hallu_rate_mean":0.6875,"hallu_prevalence":0.75,"hallu_top100":82,"inter_rater_agreement":0.7921626984126984,"n_judges":4,"tokens_in_total":11395,"tokens_out_total":20134,"model_cost_usd":0.560325},{"model_id":"claude-opus-4-8","domain":"steuerrecht","score_mean":0.6479695767195767,"score_ci_lo":0.5718650793650794,"score_ci_hi":0.7179761904761904,"top100":65,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.6479695767195767,"hallu_rate_mean":1.8,"hallu_prevalence":1,"hallu_top100":55,"inter_rater_agreement":0.6982804232804233,"n_judges":4,"tokens_in_total":9735,"tokens_out_total":20509,"model_cost_usd":0.5614},{"model_id":"claude-opus-4-8","domain":"wirtschaftsrecht","score_mean":0.6646064814814815,"score_ci_lo":0.6049735449735449,"score_ci_hi":0.727850529100529,"top100":66,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.6646064814814815,"hallu_rate_mean":1.7833333333333334,"hallu_prevalence":1,"hallu_top100":55,"inter_rater_agreement":0.6646957671957672,"n_judges":4,"tokens_in_total":22112,"tokens_out_total":45490,"model_cost_usd":1.24781},{"model_id":"gemini-3-1-pro","domain":"jura","score_mean":0.5997137188208617,"score_ci_lo":0.5239597505668935,"score_ci_hi":0.6670436507936508,"top100":60,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.5997137188208617,"hallu_rate_mean":0.7857142857142857,"hallu_prevalence":0.6,"hallu_top100":80,"inter_rater_agreement":0.6296938775510205,"n_judges":4,"tokens_in_total":12898,"tokens_out_total":6563,"model_cost_usd":0.104552},{"model_id":"gemini-3-1-pro","domain":"medizin","score_mean":0.626264880952381,"score_ci_lo":0.560689484126984,"score_ci_hi":0.6880456349206349,"top100":63,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.626264880952381,"hallu_rate_mean":0.5125,"hallu_prevalence":0.6,"hallu_top100":87,"inter_rater_agreement":0.7407738095238094,"n_judges":4,"tokens_in_total":6530,"tokens_out_total":4313,"model_cost_usd":0.064816},{"model_id":"gemini-3-1-pro","domain":"steuerrecht","score_mean":0.43675925925925924,"score_ci_lo":0.3606018518518519,"score_ci_hi":0.5068518518518519,"top100":44,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.43675925925925924,"hallu_rate_mean":1.5666666666666667,"hallu_prevalence":0.8,"hallu_top100":60,"inter_rater_agreement":0.6992063492063492,"n_judges":4,"tokens_in_total":5420,"tokens_out_total":3274,"model_cost_usd":0.050128},{"model_id":"gemini-3-1-pro","domain":"wirtschaftsrecht","score_mean":0.46411044973544974,"score_ci_lo":0.3816732804232804,"score_ci_hi":0.5459656084656085,"top100":46,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.46411044973544974,"hallu_rate_mean":1.725,"hallu_prevalence":0.9,"hallu_top100":56,"inter_rater_agreement":0.5365079365079365,"n_judges":4,"tokens_in_total":11545,"tokens_out_total":9369,"model_cost_usd":0.135518},{"model_id":"glm-4.6","domain":"jura","score_mean":0.4527210884353741,"score_ci_lo":0.37813775510204084,"score_ci_hi":0.520436507936508,"top100":45,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.4527210884353741,"hallu_rate_mean":1.2357142857142858,"hallu_prevalence":0.7714285714285715,"hallu_top100":69,"inter_rater_agreement":0.6484240362811792,"n_judges":4,"tokens_in_total":13010,"tokens_out_total":82144,"model_cost_usd":0.188523},{"model_id":"glm-4.6","domain":"medizin","score_mean":0.6033234126984127,"score_ci_lo":0.5369295634920636,"score_ci_hi":0.6683779761904762,"top100":60,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.6033234126984127,"hallu_rate_mean":0.3375,"hallu_prevalence":0.6,"hallu_top100":91,"inter_rater_agreement":0.7734126984126984,"n_judges":4,"tokens_in_total":6798,"tokens_out_total":47279,"model_cost_usd":0.108093},{"model_id":"glm-4.6","domain":"steuerrecht","score_mean":0.2323611111111111,"score_ci_lo":0.15560185185185185,"score_ci_hi":0.3147685185185185,"top100":23,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.2323611111111111,"hallu_rate_mean":2.7666666666666666,"hallu_prevalence":0.9333333333333333,"hallu_top100":30,"inter_rater_agreement":0.6231746031746032,"n_judges":4,"tokens_in_total":5575,"tokens_out_total":52395,"model_cost_usd":0.118614},{"model_id":"glm-4.6","domain":"wirtschaftsrecht","score_mean":0.44489748677248675,"score_ci_lo":0.38768849206349204,"score_ci_hi":0.4946428571428571,"top100":44,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.44489748677248675,"hallu_rate_mean":1.6166666666666667,"hallu_prevalence":0.9,"hallu_top100":59,"inter_rater_agreement":0.5324074074074073,"n_judges":4,"tokens_in_total":12340,"tokens_out_total":93763,"model_cost_usd":0.213683},{"model_id":"gpt-5","domain":"jura","score_mean":0.7279790249433107,"score_ci_lo":0.653078231292517,"score_ci_hi":0.7979421768707483,"top100":73,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.7279790249433107,"hallu_rate_mean":1.1142857142857143,"hallu_prevalence":0.8571428571428571,"hallu_top100":72,"inter_rater_agreement":0.7110430839002267,"n_judges":4,"tokens_in_total":11850,"tokens_out_total":59576,"model_cost_usd":0.95289},{"model_id":"gpt-5","domain":"medizin","score_mean":0.8425595238095237,"score_ci_lo":0.7509672619047618,"score_ci_hi":0.9154761904761906,"top100":84,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.8425595238095237,"hallu_rate_mean":0.85,"hallu_prevalence":0.9,"hallu_top100":78,"inter_rater_agreement":0.7764880952380953,"n_judges":4,"tokens_in_total":6251,"tokens_out_total":26705,"model_cost_usd":0.43183},{"model_id":"gpt-5","domain":"steuerrecht","score_mean":0.6536111111111111,"score_ci_lo":0.5493518518518518,"score_ci_hi":0.7614351851851852,"top100":65,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.6536111111111111,"hallu_rate_mean":1.8666666666666667,"hallu_prevalence":0.9333333333333333,"hallu_top100":53,"inter_rater_agreement":0.6598148148148149,"n_judges":4,"tokens_in_total":5093,"tokens_out_total":37227,"model_cost_usd":0.58387},{"model_id":"gpt-5","domain":"wirtschaftsrecht","score_mean":0.842850529100529,"score_ci_lo":0.8052314814814815,"score_ci_hi":0.8808664021164021,"top100":84,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.842850529100529,"hallu_rate_mean":1.0583333333333333,"hallu_prevalence":1,"hallu_top100":73,"inter_rater_agreement":0.7960185185185186,"n_judges":4,"tokens_in_total":11129,"tokens_out_total":65152,"model_cost_usd":1.032925},{"model_id":"gpt-5.6-sol","domain":"jura","score_mean":0.7474886621315193,"score_ci_lo":0.6949178004535147,"score_ci_hi":0.8020238095238095,"top100":75,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.7474886621315193,"hallu_rate_mean":0.9785714285714285,"hallu_prevalence":0.8571428571428571,"hallu_top100":75,"inter_rater_agreement":0.6860997732426304,"n_judges":4,"tokens_in_total":11850,"tokens_out_total":136613,"model_cost_usd":4.15764},{"model_id":"gpt-5.6-sol","domain":"medizin","score_mean":0.8482638888888889,"score_ci_lo":0.8038442460317461,"score_ci_hi":0.8921626984126985,"top100":85,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.8482638888888889,"hallu_rate_mean":0.8375,"hallu_prevalence":0.9,"hallu_top100":79,"inter_rater_agreement":0.8023809523809524,"n_judges":4,"tokens_in_total":6251,"tokens_out_total":34756,"model_cost_usd":1.073935},{"model_id":"gpt-5.6-sol","domain":"steuerrecht","score_mean":0.7049152236652237,"score_ci_lo":0.5238365800865801,"score_ci_hi":0.8496212121212121,"top100":70,"n_cells":11,"closed_accuracy":null,"open_rubric_score":0.7049152236652237,"hallu_rate_mean":2.159090909090909,"hallu_prevalence":1,"hallu_top100":46,"inter_rater_agreement":0.7374458874458875,"n_judges":4,"tokens_in_total":3535,"tokens_out_total":76383,"model_cost_usd":2.309165},{"model_id":"gpt-5.6-sol","domain":"wirtschaftsrecht","score_mean":0.7768816137566137,"score_ci_lo":0.7136044973544974,"score_ci_hi":0.8334920634920634,"top100":78,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.7768816137566137,"hallu_rate_mean":1.325,"hallu_prevalence":0.9333333333333333,"hallu_top100":66,"inter_rater_agreement":0.7081349206349207,"n_judges":4,"tokens_in_total":11129,"tokens_out_total":184744,"model_cost_usd":5.597965},{"model_id":"grok-4.5","domain":"jura","score_mean":0.5790306122448979,"score_ci_lo":0.4970039682539682,"score_ci_hi":0.6501190476190476,"top100":58,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.5790306122448979,"hallu_rate_mean":0.9,"hallu_prevalence":0.6857142857142857,"hallu_top100":77,"inter_rater_agreement":0.6627097505668934,"n_judges":4,"tokens_in_total":18826,"tokens_out_total":6127,"model_cost_usd":0.074414},{"model_id":"grok-4.5","domain":"medizin","score_mean":0.6433531746031746,"score_ci_lo":0.5565476190476191,"score_ci_hi":0.7279017857142857,"top100":64,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.6433531746031746,"hallu_rate_mean":0.475,"hallu_prevalence":0.75,"hallu_top100":88,"inter_rater_agreement":0.6898809523809524,"n_judges":4,"tokens_in_total":10172,"tokens_out_total":4903,"model_cost_usd":0.049762},{"model_id":"grok-4.5","domain":"steuerrecht","score_mean":0.4911772486772487,"score_ci_lo":0.42574074074074075,"score_ci_hi":0.561494708994709,"top100":49,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.4911772486772487,"hallu_rate_mean":1.2833333333333334,"hallu_prevalence":0.8666666666666667,"hallu_top100":67,"inter_rater_agreement":0.5685449735449736,"n_judges":4,"tokens_in_total":7983,"tokens_out_total":4488,"model_cost_usd":0.042894},{"model_id":"grok-4.5","domain":"wirtschaftsrecht","score_mean":0.5775297619047619,"score_ci_lo":0.5012301587301587,"score_ci_hi":0.6440674603174603,"top100":58,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.5775297619047619,"hallu_rate_mean":1.2916666666666667,"hallu_prevalence":0.8333333333333334,"hallu_top100":67,"inter_rater_agreement":0.621441798941799,"n_judges":4,"tokens_in_total":16875,"tokens_out_total":11577,"model_cost_usd":0.103212},{"model_id":"kimi-k2.6","domain":"jura","score_mean":0.5830286281179138,"score_ci_lo":0.5003649376417234,"score_ci_hi":0.6578550170068027,"top100":58,"n_cells":28,"closed_accuracy":null,"open_rubric_score":0.5830286281179138,"hallu_rate_mean":1.6071428571428572,"hallu_prevalence":0.7857142857142857,"hallu_top100":59,"inter_rater_agreement":0.6420351473922903,"n_judges":4,"tokens_in_total":12008,"tokens_out_total":165930,"model_cost_usd":0.675128},{"model_id":"kimi-k2.6","domain":"medizin","score_mean":0.7134672619047618,"score_ci_lo":0.6297619047619047,"score_ci_hi":0.7891865079365079,"top100":71,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.7134672619047618,"hallu_rate_mean":0.9125,"hallu_prevalence":0.95,"hallu_top100":77,"inter_rater_agreement":0.7148809523809524,"n_judges":4,"tokens_in_total":7536,"tokens_out_total":93706,"model_cost_usd":0.381983},{"model_id":"kimi-k2.6","domain":"steuerrecht","score_mean":0.3548611111111111,"score_ci_lo":0.206875,"score_ci_hi":0.4892361111111111,"top100":35,"n_cells":10,"closed_accuracy":null,"open_rubric_score":0.3548611111111111,"hallu_rate_mean":3.375,"hallu_prevalence":1,"hallu_top100":15,"inter_rater_agreement":0.6158730158730159,"n_judges":4,"tokens_in_total":4071,"tokens_out_total":84125,"model_cost_usd":0.340367},{"model_id":"kimi-k2.6","domain":"wirtschaftsrecht","score_mean":0.643060064935065,"score_ci_lo":0.6018939393939394,"score_ci_hi":0.6842667748917749,"top100":64,"n_cells":22,"closed_accuracy":null,"open_rubric_score":0.643060064935065,"hallu_rate_mean":2.4886363636363638,"hallu_prevalence":1,"hallu_top100":37,"inter_rater_agreement":0.5846320346320346,"n_judges":4,"tokens_in_total":10071,"tokens_out_total":171663,"model_cost_usd":0.696219},{"model_id":"mistral-large-2","domain":"jura","score_mean":0.4516184807256236,"score_ci_lo":0.37185657596371885,"score_ci_hi":0.5401785714285714,"top100":45,"n_cells":35,"closed_accuracy":null,"open_rubric_score":0.4516184807256236,"hallu_rate_mean":1.9142857142857144,"hallu_prevalence":0.7714285714285715,"hallu_top100":52,"inter_rater_agreement":0.6588208616780045,"n_judges":4,"tokens_in_total":13332,"tokens_out_total":10317,"model_cost_usd":0.088566},{"model_id":"mistral-large-2","domain":"medizin","score_mean":0.6563244047619048,"score_ci_lo":0.6020833333333333,"score_ci_hi":0.711235119047619,"top100":66,"n_cells":20,"closed_accuracy":null,"open_rubric_score":0.6563244047619048,"hallu_rate_mean":1.0125,"hallu_prevalence":0.85,"hallu_top100":74,"inter_rater_agreement":0.6294642857142857,"n_judges":4,"tokens_in_total":6669,"tokens_out_total":10146,"model_cost_usd":0.074214},{"model_id":"mistral-large-2","domain":"steuerrecht","score_mean":0.24302910052910054,"score_ci_lo":0.15027777777777776,"score_ci_hi":0.3488888888888889,"top100":24,"n_cells":15,"closed_accuracy":null,"open_rubric_score":0.24302910052910054,"hallu_rate_mean":3.966666666666667,"hallu_prevalence":1,"hallu_top100":0,"inter_rater_agreement":0.4753439153439153,"n_judges":4,"tokens_in_total":5572,"tokens_out_total":7315,"model_cost_usd":0.055034},{"model_id":"mistral-large-2","domain":"wirtschaftsrecht","score_mean":0.5710185185185186,"score_ci_lo":0.4972486772486773,"score_ci_hi":0.6422685185185185,"top100":57,"n_cells":30,"closed_accuracy":null,"open_rubric_score":0.5710185185185186,"hallu_rate_mean":2.025,"hallu_prevalence":0.9666666666666667,"hallu_top100":49,"inter_rater_agreement":0.5450132275132274,"n_judges":4,"tokens_in_total":11870,"tokens_out_total":18555,"model_cost_usd":0.13507}],"by_model_overall":[{"model_id":"claude-opus-4-7","score_mean_across_domains":0.7938571428571428,"score_ci_lo":0.7631388888888888,"score_ci_hi":0.8202906746031746,"top100":79,"n_cells":100,"hallu_rate_mean_across_domains":1.625,"hallu_top100_across_domains":59},{"model_id":"claude-opus-4-8","score_mean_across_domains":0.7395525793650793,"score_ci_lo":0.7086547619047618,"score_ci_hi":0.769797619047619,"top100":74,"n_cells":100,"hallu_rate_mean_across_domains":1.175,"hallu_top100_across_domains":70},{"model_id":"gemini-3-1-pro","score_mean_across_domains":0.5398998015873016,"score_ci_lo":0.4936140873015873,"score_ci_hi":0.5821458333333333,"top100":54,"n_cells":100,"hallu_rate_mean_across_domains":1.13,"hallu_top100_across_domains":71},{"model_id":"glm-4.6","score_mean_across_domains":0.4474404761904762,"score_ci_lo":0.4028988095238095,"score_ci_hi":0.4839196428571429,"top100":45,"n_cells":100,"hallu_rate_mean_across_domains":1.4,"hallu_top100_across_domains":65},{"model_id":"gpt-5","score_mean_across_domains":0.7742013888888889,"score_ci_lo":0.7377450396825397,"score_ci_hi":0.8104047619047619,"top100":77,"n_cells":100,"hallu_rate_mean_across_domains":1.1575,"hallu_top100_across_domains":71},{"model_id":"gpt-5.6-sol","score_mean_across_domains":0.7727905919312169,"score_ci_lo":0.7326409556878306,"score_ci_hi":0.8078124999999999,"top100":77,"n_cells":96,"hallu_rate_mean_across_domains":1.1927083333333333,"hallu_top100_across_domains":70},{"model_id":"grok-4.5","score_mean_across_domains":0.5782668650793651,"score_ci_lo":0.5366686507936508,"score_ci_hi":0.6172321428571429,"top100":58,"n_cells":100,"hallu_rate_mean_across_domains":0.99,"hallu_top100_across_domains":75},{"model_id":"kimi-k2.6","score_mean_across_domains":0.6036259920634921,"score_ci_lo":0.5556622023809524,"score_ci_hi":0.6519022817460318,"top100":60,"n_cells":80,"hallu_rate_mean_across_domains":1.896875,"hallu_top100_across_domains":52},{"model_id":"mistral-large-2","score_mean_across_domains":0.49709126984126983,"score_ci_lo":0.4494047619047619,"score_ci_hi":0.5428849206349207,"top100":50,"n_cells":100,"hallu_rate_mean_across_domains":2.075,"hallu_top100_across_domains":48}],"oracle_by_domain":[{"domain":"jura","n_items":35,"oracle_mean":0.8784,"oracle_top100":88,"best_single_model_id":"claude-opus-4-7","best_single_model_mean":0.8086,"best_single_model_top100":81},{"domain":"medizin","n_items":20,"oracle_mean":0.9251,"oracle_top100":93,"best_single_model_id":"gpt-5.6-sol","best_single_model_mean":0.8483,"best_single_model_top100":85},{"domain":"steuerrecht","n_items":15,"oracle_mean":0.8364,"oracle_top100":84,"best_single_model_id":"claude-opus-4-7","best_single_model_mean":0.7216,"best_single_model_top100":72},{"domain":"wirtschaftsrecht","n_items":30,"oracle_mean":0.8783,"oracle_top100":88,"best_single_model_id":"gpt-5","best_single_model_mean":0.8428,"best_single_model_top100":84}],"cost":{"model_calls_usd":26.294,"judge_calls_usd":91.6231,"total_usd":117.9171,"by_judge":[{"judge_id":"claude-opus-4-7","tokens_in_total":2443222,"tokens_out_total":534912,"cost_usd":25.58891},{"judge_id":"claude-opus-4-8","tokens_in_total":2438842,"tokens_out_total":687411,"cost_usd":29.379485},{"judge_id":"gpt-5","tokens_in_total":1258888,"tokens_out_total":1610844,"cost_usd":30.4571},{"judge_id":"mistral-large-2","tokens_in_total":1350566,"tokens_out_total":582739,"cost_usd":6.197566}],"extrapolation":{"scale_factor":12,"extrapolated_total_usd":1415,"target_full_run_cells":1200,"observed_cells":100,"note":"linear extrapolation; assumes uniform prompt sizes + judge mix"},"missing_pricing_for":[]}},"raw_jsonl_url":"https://licenses.ai-roundtable.de/r2/runs/run-20260715T040307-80bf67/raw.jsonl","progress_jsonl_url":"https://licenses.ai-roundtable.de/r2/runs/run-20260715T040307-80bf67/progress.jsonl","cost_total_usd":117.9171}