tokenusage-cli 0.4.3 → 0.4.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -556,6 +556,597 @@
556
556
  "cache_read": 0,
557
557
  "cache_write": 0,
558
558
  "note": "Alias for free Build SKU as labeled in Grok updates.jsonl modelUsage. Canonical storage id is grok-build-free."
559
+ },
560
+ "claude-opus-4-5": {
561
+ "input": 5,
562
+ "output": 25,
563
+ "cache_read": 0.5,
564
+ "cache_write": 6.25,
565
+ "note": "Canonical undated Opus 4.5 alias (LiteLLM bare key, sourced from https://platform.claude.com/docs/en/about\u002dclaude/pricing, cached 2026\u002d10\u002d31): standard Opus tier $5/$25 per MTok, cache_read 0.1x, cache_write 1.25x — same rates as claude\u002dopus\u002d4\u002d5\u002d20250414. Previously LiteLLM\u002donly; promoted so cloud account views price the bare id (issue #929)."
566
+ },
567
+ "claude-opus-4-6": {
568
+ "input": 5,
569
+ "output": 25,
570
+ "cache_read": 0.5,
571
+ "cache_write": 6.25,
572
+ "note": "Opus 4.6 standard tier (LiteLLM entry sourced from https://platform.claude.com/docs/en/about\u002dclaude/pricing, cached 2026\u002d10\u002d31): $5/$25 per MTok, cache_read 0.1x, cache_write 1.25x. The edge table pins the same rate; curated pins it so local billing and \u002dthinking variants no longer depend on LiteLLM fuzzy matching (issue #929)."
573
+ },
574
+ "claude-sonnet-4-6": {
575
+ "input": 3,
576
+ "output": 15,
577
+ "cache_read": 0.3,
578
+ "cache_write": 3.75,
579
+ "note": "Sonnet 4.6 standard tier (LiteLLM entry sourced from https://platform.claude.com/docs/en/about\u002dclaude/pricing, cached 2026\u002d10\u002d31): $3/$15 per MTok, cache_read 0.1x, cache_write 1.25x. The edge table pins the same rate; curated pins it so local billing no longer depends on the LiteLLM cache (issue #929)."
580
+ },
581
+ "claude-haiku-5-5": {
582
+ "input": 0.1,
583
+ "output": 0.5,
584
+ "cache_read": 0.01,
585
+ "cache_write": 0.125,
586
+ "note": "Official Anthropic Haiku 5.5 short\u002dcontext (prompts <=100K) USD/MTok verified 2026\u002d10\u002d22: https://platform.claude.com/docs/en/about\u002dclaude/pricing. Prompts over 100K use 5x rates; computeRowCost does not yet split that long\u002dprompt tier. Cache write is the 5\u002dminute rate."
587
+ },
588
+ "claude-haiku-5.5": {
589
+ "input": 0.1,
590
+ "output": 0.5,
591
+ "cache_read": 0.01,
592
+ "cache_write": 0.125,
593
+ "note": "Dotted alias for claude\u002dhaiku\u002d5\u002d5. Same short\u002dcontext rates."
594
+ },
595
+ "gpt-6.1-terra": {
596
+ "input": 1,
597
+ "output": 5,
598
+ "cache_read": 0.05,
599
+ "cache_write": 1.25,
600
+ "priority_multiplier": 2,
601
+ "note": "Official Standard USD/MTok rates verified 2026\u002d10\u002d27: https://developers.openai.com/api/docs/models/gpt\u002d6.1\u002dterra and https://developers.openai.com/api/docs/pricing. Short\u002dcontext (≤272K): $1/$5/$0.05 cached/$1.25 cache write (30m). Requests above 272K input use 2x input/cache and 1.5x output for the full request. Fast mode is 2x Standard; Ultrafast is not published for this model. Premiums apply only to observed request subsets. Cloud aggregates remain Standard short\u002dcontext estimates."
602
+ },
603
+ "gpt-6-luna": {
604
+ "input": 0.1,
605
+ "output": 0.5,
606
+ "cache_read": 0.01,
607
+ "cache_write": 0.125,
608
+ "note": "GPT\u002d6 Luna lightweight tier, Standard USD/MTok (LiteLLM entry sourced from https://developers.openai.com/api/docs/pricing, cached 2026\u002d10\u002d31): $0.10/$0.50 per MTok, $0.01 cache read, $0.125 cache write. Distinct from gpt\u002d5.6\u002dluna. Previously LiteLLM\u002donly, so cloud account views showed it unpriced; promoted so cloud prices it and curated/edge parity guards the key (issue #929)."
609
+ },
610
+ "gpt-5.5": {
611
+ "input": 5,
612
+ "output": 30,
613
+ "cache_read": 0.5,
614
+ "note": "Official Standard USD/MTok rates (LiteLLM entry sourced from https://developers.openai.com/api/docs/pricing, cached 2026\u002d10\u002d31): $5/$30 per MTok, $0.50 cache read, no published cache\u002dwrite rate. The largest production LiteLLM\u002donly model — codex's primary SKU ($1,118 of cloud account usage showed unpriced between 2026\u002d01\u002d01 and 2026\u002d10\u002d31); promoted so cloud account views price it and curated/edge parity guards the key (issue #929)."
615
+ },
616
+ "gpt-5.4": {
617
+ "input": 2.5,
618
+ "output": 15,
619
+ "cache_read": 0.25,
620
+ "note": "Official Standard USD/MTok rates (LiteLLM entry sourced from https://developers.openai.com/api/docs/pricing, cached 2026\u002d10\u002d31): $2.50/$15 per MTok, $0.25 cache read. The edge table has pinned this rate since before v1.5.1; curated pins the same value so local billing no longer depends on the LiteLLM cache (issue #929). \u002dpro/\u002dnano/\u002dmini keep their own SKUs."
621
+ },
622
+ "gpt-5.4-pro": {
623
+ "input": 30,
624
+ "output": 180,
625
+ "cache_read": 3,
626
+ "note": "Official Pro USD/MTok rates, matching the edge table's gpt\u002d5.4\u002dpro entry ($30/$180, $3 cache read; https://developers.openai.com/api/docs/pricing). Companion pin for the gpt\u002d5.4 promotion: without a curated Pro entry, the curated contained match captures provider\u002dqualified dated snapshots (openai/gpt\u002d5.4\u002dpro\u002d2026\u002d03\u002d05) at the base $2.50/$15 rate before the LiteLLM reverse lookup reaches the Pro SKU — 12x under local vs edge (issue #929)."
627
+ },
628
+ "gpt-5.4-mini": {
629
+ "input": 0.75,
630
+ "output": 4.5,
631
+ "cache_read": 0.075,
632
+ "note": "Official Standard USD/MTok rates (LiteLLM entry sourced from https://developers.openai.com/api/docs/pricing, cached 2026\u002d10\u002d31): $0.75/$4.50 per MTok, $0.075 cache read. Same promotion rationale as gpt\u002d5.4 (issue #929)."
633
+ },
634
+ "gpt-5.2": {
635
+ "input": 1.75,
636
+ "output": 14,
637
+ "cache_read": 0.175,
638
+ "note": "Official Standard USD/MTok rates (LiteLLM entry sourced from https://developers.openai.com/api/docs/pricing, cached 2026\u002d10\u002d31): $1.75/$14 per MTok, $0.175 cache read. Same promotion rationale as gpt\u002d5.4 (issue #929); \u002dpro stays on its own 10x LiteLLM SKU."
639
+ },
640
+ "gpt-5.3-codex-spark": {
641
+ "input": 1.75,
642
+ "output": 14,
643
+ "cache_read": 0.175,
644
+ "note": "Codex spark capacity tier; not a separate LiteLLM key. Local (LiteLLM reverse\u002dsubstring) and the edge matcher both bill it at the base gpt\u002d5.3\u002dcodex rate ($1.75/$14, $0.175 cache read, https://developers.openai.com/api/docs/pricing); the exact entry pins that resolution on both sides (issue #929)."
645
+ },
646
+ "qwen3.8-max": {
647
+ "input": 2,
648
+ "output": 6,
649
+ "cache_read": 0.17,
650
+ "cache_write": 2.5,
651
+ "note": "Alibaba Model Studio Singapore (International), USD per MTok, verified 2026\u002d10\u002d25: https://www.alibabacloud.com/help/en/model\u002dstudio/model\u002dpricing and https://www.alibabacloud.com/help/en/model\u002dstudio/qwen3\u002d8\u002dmax . Reference API estimate; other regions/providers can have different rates. Issue #744."
652
+ },
653
+ "qwen3.8-max-0902": {
654
+ "input": 2,
655
+ "output": 6,
656
+ "cache_read": 0.17,
657
+ "cache_write": 2.5,
658
+ "note": "Alibaba Model Studio Singapore (International), USD per MTok, verified 2026\u002d10\u002d25: https://www.alibabacloud.com/help/en/model\u002dstudio/model\u002dpricing and https://www.alibabacloud.com/help/en/model\u002dstudio/qwen3\u002d8\u002dmax . Reference API estimate; other regions/providers can have different rates. Issue #744. Dated SKU qwen3.8\u002dmax\u002d0902 shares Singapore International list rates."
659
+ },
660
+ "qwen3.8-max-2026-09-02": {
661
+ "input": 2,
662
+ "output": 6,
663
+ "cache_read": 0.17,
664
+ "cache_write": 2.5,
665
+ "note": "Alibaba Model Studio Singapore (International), USD per MTok, verified 2026\u002d10\u002d25: https://www.alibabacloud.com/help/en/model\u002dstudio/qwen3\u002d8\u002dmax . Alias of qwen3.8\u002dmax\u002d0902; same Singapore International list rates. Issue #744."
666
+ },
667
+ "gemini-3-flash-preview": {
668
+ "input": 0.5,
669
+ "output": 3,
670
+ "cache_read": 0.05,
671
+ "note": "LiteLLM entry (sourced from https://cloud.google.com/gemini\u002denterprise\u002dagent\u002dplatform/generative\u002dai/pricing, cached 2026\u002d10\u002d31): $0.50/$3 per MTok, $0.05 cache read, no published cache\u002dwrite rate. Same rate the edge table already pins; promoted so curated/edge parity guards the key (issue #929)."
672
+ },
673
+ "kimi-k3.1-pro": {
674
+ "input": 2,
675
+ "output": 10,
676
+ "cache_read": 0.2,
677
+ "cache_write": 2,
678
+ "note": "Kimi K3.1 Pro (API id kimi\u002dk3.1\u002dpro). Official USD list from https://platform.kimi.ai/ : $2/M input, $10/M output, $2/M cache write (default 5m TTL), $0.20/M cache hit. Flat 1M context; no context\u002dlength tiers. Verified 2026\u002d10\u002d25. Distinct from kimi\u002dk3 ($3/$15)."
679
+ },
680
+ "kimi-k3.1-pro-highspeed": {
681
+ "input": 4,
682
+ "output": 20,
683
+ "cache_read": 0.2,
684
+ "cache_write": 4,
685
+ "note": "Kimi K3.1 Pro Highspeed (API id kimi\u002dk3.1\u002dpro\u002dhighspeed). Official USD list from https://platform.kimi.ai/ : $4/M input, $20/M output, $4/M cache write (default 5m TTL), $0.20/M cache hit. Same intelligence as K3.1 Pro at higher speed. Verified 2026\u002d10\u002d25."
686
+ },
687
+ "k3.1-pro": {
688
+ "input": 2,
689
+ "output": 10,
690
+ "cache_read": 0.2,
691
+ "cache_write": 2,
692
+ "note": "Bare/compact alias for kimi\u002dk3.1\u002dpro. Same official rates."
693
+ },
694
+ "k3.1pro": {
695
+ "input": 2,
696
+ "output": 10,
697
+ "cache_read": 0.2,
698
+ "cache_write": 2,
699
+ "note": "Compact alias (no hyphen) for kimi\u002dk3.1\u002dpro. Same official rates."
700
+ },
701
+ "k3.1-pro-highspeed": {
702
+ "input": 4,
703
+ "output": 20,
704
+ "cache_read": 0.2,
705
+ "cache_write": 4,
706
+ "note": "Bare alias for kimi\u002dk3.1\u002dpro\u002dhighspeed. Same official rates."
707
+ },
708
+ "k3.1pro-highspeed": {
709
+ "input": 4,
710
+ "output": 20,
711
+ "cache_read": 0.2,
712
+ "cache_write": 4,
713
+ "note": "Compact alias for kimi\u002dk3.1\u002dpro\u002dhighspeed. Same official rates."
714
+ },
715
+ "mimo-v2.5": {
716
+ "input": 0.4,
717
+ "output": 2,
718
+ "cache_read": 0.08,
719
+ "note": "Xiaomi MiMo v2.5 (queue rows report the bare id; LiteLLM keys are provider\u002dqualified). Rates match the deepinfra entry — the lexicographically smallest provider key the matcher deterministically picks: $0.40/$2 per MTok, $0.08 cache read, no published cache\u002dwrite rate. The edge table pinned this exact rate after the kimi\u002dk2.6/mimo profile\u002dmodal incident; curated pins the same value so local billing no longer depends on LiteLLM prefix\u002dstrip (issue #929)."
720
+ },
721
+ "mimo-v2.5-pro": {
722
+ "input": 1,
723
+ "output": 3,
724
+ "cache_read": 0.2,
725
+ "note": "Xiaomi MiMo v2.5 Pro (deepinfra is the lexicographically smallest LiteLLM provider key, matching the edge table): $1/$3 per MTok, $0.20 cache read. Exact entry required alongside mimo\u002dv2.5 so the contained\u002dexact fallback cannot bill \u002dpro at the base v2.5 rate (issue #929)."
726
+ },
727
+ "muse-spark-1.2-contributor": {
728
+ "input": 0.1,
729
+ "output": 0.2,
730
+ "cache_read": 0.002,
731
+ "note": "Paid Muse contributor SKU. LiteLLM provider\u002dqualified entries agree on $0.10/$0.20 per MTok, $0.002 cache read (meta and opencode_go both list these rates, cached 2026\u002d10\u002d31). Previously LiteLLM prefix\u002dstrip\u002donly, so cloud account views showed it unpriced; promoted so cloud prices it and curated/edge parity guards the key (issue #929). The \u002dfree SKU below stays $0."
732
+ },
733
+ "muse-spark-1.3-contributor": {
734
+ "input": 0.1,
735
+ "output": 0.2,
736
+ "cache_read": 0.002,
737
+ "note": "See muse\u002dspark\u002d1.2\u002dcontributor — same paid contributor SKU at the same verified rates (issue #929)."
738
+ },
739
+ "muse-spark-1.2-contributor-free": {
740
+ "input": 0,
741
+ "output": 0,
742
+ "cache_read": 0,
743
+ "cache_write": 0,
744
+ "note": "Muse contributor free SKU. Also covered by matcher \u002dfree suffix; exact entry keeps status=free (issue #744)."
745
+ },
746
+ "muse-spark-1.3-contributor-free": {
747
+ "input": 0,
748
+ "output": 0,
749
+ "cache_read": 0,
750
+ "cache_write": 0,
751
+ "note": "Muse contributor free SKU. Also covered by matcher \u002dfree suffix; exact entry keeps status=free (issue #744)."
752
+ },
753
+ "qvq-max": {
754
+ "input": 1.147,
755
+ "output": 4.588,
756
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
757
+ },
758
+ "qwen-deep-research": {
759
+ "input": 7.742,
760
+ "output": 23.367,
761
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
762
+ },
763
+ "qwen-doc-turbo": {
764
+ "input": 0.087,
765
+ "output": 0.144,
766
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
767
+ },
768
+ "qwen-flash": {
769
+ "input": 0.022,
770
+ "output": 0.216,
771
+ "cache_read": 0.0044,
772
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
773
+ },
774
+ "qwen-long": {
775
+ "input": 0.072,
776
+ "output": 0.287,
777
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
778
+ },
779
+ "qwen-math-plus": {
780
+ "input": 0.574,
781
+ "output": 1.721,
782
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
783
+ },
784
+ "qwen-math-turbo": {
785
+ "input": 0.287,
786
+ "output": 0.861,
787
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
788
+ },
789
+ "qwen-mt-turbo": {
790
+ "input": 0.101,
791
+ "output": 0.28,
792
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
793
+ },
794
+ "qwen-omni-turbo": {
795
+ "input": 0.058,
796
+ "output": 0.23,
797
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
798
+ },
799
+ "qwen-plus-character": {
800
+ "input": 0.115,
801
+ "output": 0.287,
802
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
803
+ },
804
+ "qwen-vl-max": {
805
+ "input": 0.23,
806
+ "output": 0.574,
807
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
808
+ },
809
+ "qwen-vl-ocr": {
810
+ "input": 0.043,
811
+ "output": 0.072,
812
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
813
+ },
814
+ "qwen2-5-14b-instruct": {
815
+ "input": 0.144,
816
+ "output": 0.431,
817
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
818
+ },
819
+ "qwen2-5-32b-instruct": {
820
+ "input": 0.287,
821
+ "output": 0.861,
822
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
823
+ },
824
+ "qwen2-5-72b-instruct": {
825
+ "input": 0.574,
826
+ "output": 1.721,
827
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
828
+ },
829
+ "qwen2-5-7b-instruct": {
830
+ "input": 0.072,
831
+ "output": 0.144,
832
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
833
+ },
834
+ "qwen2-5-coder-32b-instruct": {
835
+ "input": 0.287,
836
+ "output": 0.861,
837
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
838
+ },
839
+ "qwen2-5-coder-7b-instruct": {
840
+ "input": 0.144,
841
+ "output": 0.287,
842
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
843
+ },
844
+ "qwen2-5-math-72b-instruct": {
845
+ "input": 0.574,
846
+ "output": 1.721,
847
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
848
+ },
849
+ "qwen2-5-math-7b-instruct": {
850
+ "input": 0.144,
851
+ "output": 0.287,
852
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
853
+ },
854
+ "qwen2-5-omni-7b": {
855
+ "input": 0.087,
856
+ "output": 0.345,
857
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
858
+ },
859
+ "qwen2-5-vl-72b-instruct": {
860
+ "input": 2.294,
861
+ "output": 6.881,
862
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
863
+ },
864
+ "qwen2-5-vl-7b-instruct": {
865
+ "input": 0.287,
866
+ "output": 0.717,
867
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
868
+ },
869
+ "qwen3-asr-flash": {
870
+ "input": 0.032,
871
+ "output": 0.032,
872
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
873
+ },
874
+ "qwen3-coder-flash": {
875
+ "input": 0.144,
876
+ "output": 0.574,
877
+ "cache_read": 0.0288,
878
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
879
+ },
880
+ "qwen3-omni-flash": {
881
+ "input": 0.258,
882
+ "output": 0.989,
883
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
884
+ },
885
+ "qwen3-vl-235b-a22b": {
886
+ "input": 0.286705,
887
+ "output": 1.14682,
888
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
889
+ },
890
+ "qwen3-vl-30b-a3b": {
891
+ "input": 0.108,
892
+ "output": 0.431,
893
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
894
+ },
895
+ "qwen3-vl-plus": {
896
+ "input": 0.143353,
897
+ "output": 1.433525,
898
+ "cache_read": 0.028671,
899
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
900
+ },
901
+ "qwen3.5-flash": {
902
+ "input": 0.029,
903
+ "output": 0.287,
904
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
905
+ },
906
+ "qwen3.5-plus": {
907
+ "input": 0.115,
908
+ "output": 0.688,
909
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
910
+ },
911
+ "qwen3.6-max-preview": {
912
+ "input": 1.32,
913
+ "output": 7.9,
914
+ "cache_read": 0.132,
915
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
916
+ },
917
+ "qwen3.7-flash": {
918
+ "input": 0.02962,
919
+ "output": 0.1185,
920
+ "cache_read": 0.005924,
921
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
922
+ },
923
+ "qwen3.7-plus": {
924
+ "input": 0.276,
925
+ "output": 1.101,
926
+ "cache_read": 0.0552,
927
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
928
+ },
929
+ "tongyi-intent-detect-v3": {
930
+ "input": 0.058,
931
+ "output": 0.144,
932
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
933
+ },
934
+ "doubao-lite-128k": {
935
+ "input": 0.041,
936
+ "output": 0.083,
937
+ "cache_read": 0.008,
938
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
939
+ },
940
+ "doubao-pro-128k": {
941
+ "input": 0.111,
942
+ "output": 0.277,
943
+ "cache_read": 0.022,
944
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
945
+ },
946
+ "doubao-seed-2.0-code": {
947
+ "input": 0.449,
948
+ "output": 2.282,
949
+ "cache_read": 0.09,
950
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
951
+ },
952
+ "doubao-seed-2.0-lite": {
953
+ "input": 0.083,
954
+ "output": 0.498,
955
+ "cache_read": 0.017,
956
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
957
+ },
958
+ "doubao-seed-2.0-mini": {
959
+ "input": 0.028,
960
+ "output": 0.277,
961
+ "cache_read": 0.006,
962
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
963
+ },
964
+ "doubao-seed-2.0-pro": {
965
+ "input": 0.442,
966
+ "output": 2.213,
967
+ "cache_read": 0.088,
968
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
969
+ },
970
+ "doubao-seed-2.1-pro": {
971
+ "input": 0.829,
972
+ "output": 4.147,
973
+ "cache_read": 0.166,
974
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
975
+ },
976
+ "doubao-seed-2.1-turbo": {
977
+ "input": 0.415,
978
+ "output": 2.073,
979
+ "cache_read": 0.083,
980
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
981
+ },
982
+ "doubao-seed-code": {
983
+ "input": 0.166,
984
+ "output": 1.106,
985
+ "cache_read": 0.033,
986
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
987
+ },
988
+ "doubao-seed-evolving": {
989
+ "input": 0.8906,
990
+ "output": 4.45301,
991
+ "cache_read": 0.17812,
992
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
993
+ },
994
+ "mimo-v2-omni": {
995
+ "input": 0.14,
996
+ "output": 0.28,
997
+ "cache_read": 0.0028,
998
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
999
+ },
1000
+ "mimo-v2-pro": {
1001
+ "input": 0.435,
1002
+ "output": 0.87,
1003
+ "cache_read": 0.0036,
1004
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1005
+ },
1006
+ "mimo-v2.5-pro-ultraspeed": {
1007
+ "input": 1.305,
1008
+ "output": 2.61,
1009
+ "cache_read": 0.0108,
1010
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1011
+ },
1012
+ "mimo-v2.6-flash": {
1013
+ "input": 0.14,
1014
+ "output": 0.28,
1015
+ "cache_read": 0.0028,
1016
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1017
+ },
1018
+ "mimo-v2.6-pro": {
1019
+ "input": 0.435,
1020
+ "output": 0.87,
1021
+ "cache_read": 0.0036,
1022
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1023
+ },
1024
+ "mimo-v2.6-pro-ultraspeed": {
1025
+ "input": 4.35,
1026
+ "output": 8.7,
1027
+ "cache_read": 0.036,
1028
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1029
+ },
1030
+ "muse-spark-1.1": {
1031
+ "input": 1.25,
1032
+ "output": 4.25,
1033
+ "cache_read": 0.15,
1034
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1035
+ },
1036
+ "muse-spark-1.2": {
1037
+ "input": 1.25,
1038
+ "output": 4.25,
1039
+ "cache_read": 0.15,
1040
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1041
+ },
1042
+ "qwen3.7-max": {
1043
+ "input": 1.65,
1044
+ "output": 4.951,
1045
+ "cache_read": 0.33,
1046
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1047
+ },
1048
+ "qwen3.6-flash": {
1049
+ "input": 0.165,
1050
+ "output": 0.99,
1051
+ "cache_read": 0.0165,
1052
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1053
+ },
1054
+ "doubao-seed-2-0-pro-260215": {
1055
+ "input": 0.47499,
1056
+ "output": 2.37494,
1057
+ "cache_read": 0.095,
1058
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1059
+ },
1060
+ "doubao-seed-1-6-251015": {
1061
+ "input": 0.11875,
1062
+ "output": 1.18747,
1063
+ "cache_read": 0.02375,
1064
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1065
+ },
1066
+ "doubao-seed-2-1-pro-260628": {
1067
+ "input": 0.8906,
1068
+ "output": 4.45301,
1069
+ "cache_read": 0.17812,
1070
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1071
+ },
1072
+ "doubao-seed-1-6-flash-250828": {
1073
+ "input": 0.02227,
1074
+ "output": 0.22265,
1075
+ "cache_read": 0.00445,
1076
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1077
+ },
1078
+ "doubao-seed-2-0-lite-260428": {
1079
+ "input": 0.08906,
1080
+ "output": 0.53436,
1081
+ "cache_read": 0.01781,
1082
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1083
+ },
1084
+ "doubao-seed-2-0-code-preview-260215": {
1085
+ "input": 0.47499,
1086
+ "output": 2.37494,
1087
+ "cache_read": 0.095,
1088
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1089
+ },
1090
+ "doubao-seed-1-6-vision-250815": {
1091
+ "input": 0.11875,
1092
+ "output": 1.18747,
1093
+ "cache_read": 0.02375,
1094
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1095
+ },
1096
+ "doubao-seed-2-0-mini-260428": {
1097
+ "input": 0.02969,
1098
+ "output": 0.29687,
1099
+ "cache_read": 0.00594,
1100
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1101
+ },
1102
+ "doubao-seed-2-1-turbo-260628": {
1103
+ "input": 0.4453,
1104
+ "output": 2.22651,
1105
+ "cache_read": 0.08906,
1106
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1107
+ },
1108
+ "doubao-seed-character-260628": {
1109
+ "input": 0.11875,
1110
+ "output": 0.29687,
1111
+ "cache_read": 0.02375,
1112
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1113
+ },
1114
+ "doubao-seed-1-8-251228": {
1115
+ "input": 0.11875,
1116
+ "output": 1.18747,
1117
+ "cache_read": 0.02375,
1118
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1119
+ },
1120
+ "muse-spark-1.3": {
1121
+ "input": 1.25,
1122
+ "output": 4.25,
1123
+ "cache_read": 0.15,
1124
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1125
+ },
1126
+ "qwen3.8-omni-flash": {
1127
+ "input": 0.113,
1128
+ "output": 0.382,
1129
+ "cache_read": 0.014,
1130
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1131
+ },
1132
+ "muse-wonder-1.0": {
1133
+ "input": 4,
1134
+ "output": 20,
1135
+ "cache_read": 0.2,
1136
+ "note": "Imported from juejin\u002dusage@f72ca4d packages/core/src/pricing/pricing.json (models.dev official\u002dchannel bundle, MIT (c) juejin\u002dcn) on 2026\u002d11\u002d03. No published cache_write rate."
1137
+ },
1138
+ "u2-flash": {
1139
+ "input": 0.14,
1140
+ "output": 0.28,
1141
+ "cache_read": 0.03,
1142
+ "note": "Unisound U2-Flash (maas.unisound.com), pay-as-you-go tier verified 2026-11-03: https://maas.unisound.com/docs/price/overview \u2014 CNY 1 / 2 per MTok input/output, CNY 0.2 cache hit, converted at 7.2 CNY/USD and rounded to 2dp (same convention as the iFlytek entries). No published cache-write rate. Collected via the DeepSeek Harness passive reader (source dsh). If usage actually rides a U Plan subscription this reference rate overstates real spend \u2014 the same estimation convention as other subscription-with-local-tokens sources (grok/kimi/zcode)."
1143
+ },
1144
+ "step-5-preview": {
1145
+ "input": 1,
1146
+ "output": 2.7,
1147
+ "cache_read": 0.1,
1148
+ "cache_write": 1,
1149
+ "note": "StepFun Step 5 Preview, official USD list verified 2026-11-03: https://platform.stepfun.ai/docs/en/guides/pricing/details ($1.00 / $2.70 per MTok input/output, $0.10 cache hit). The China site lists CNY 7 / 20 with a CNY 0.70 cache-hit rate (raised from CNY 0.35 on 2026-10-14): https://platform.stepfun.com/docs/zh/guides/pricing/details. Output includes the reasoning trace (no separate reasoning rate); the cache-miss input price includes writing new content to the cache, so cache_write = input. Promoted from selfhost-overrides on 2026-11-03 (first verified there 2026-10-22). Step Plan subscription usage would make this reference rate an overestimate \u2014 same convention as other subscription-with-local-tokens sources (grok/kimi/zcode). OpenCode's own free SKU is the suffixed step-5-preview-free and stays zero via the opencode -free rule."
559
1150
  }
560
1151
  },
561
1152
  "alias": {
@@ -817,6 +1408,42 @@
817
1408
  {
818
1409
  "match": "stepfun",
819
1410
  "ref": "step-3.7-flash"
1411
+ },
1412
+ {
1413
+ "match": "gpt\u002d6.1\u002dterra",
1414
+ "ref": "gpt\u002d6.1\u002dterra"
1415
+ },
1416
+ {
1417
+ "match": "haiku\u002d5\u002d5",
1418
+ "ref": "claude\u002dhaiku\u002d5\u002d5"
1419
+ },
1420
+ {
1421
+ "match": "haiku\u002d5.5",
1422
+ "ref": "claude\u002dhaiku\u002d5.5"
1423
+ },
1424
+ {
1425
+ "match": "kimi\u002dk3.1\u002dpro\u002dhighspeed",
1426
+ "ref": "kimi\u002dk3.1\u002dpro\u002dhighspeed"
1427
+ },
1428
+ {
1429
+ "match": "k3.1\u002dpro\u002dhighspeed",
1430
+ "ref": "kimi\u002dk3.1\u002dpro\u002dhighspeed"
1431
+ },
1432
+ {
1433
+ "match": "k3.1pro\u002dhighspeed",
1434
+ "ref": "kimi\u002dk3.1\u002dpro\u002dhighspeed"
1435
+ },
1436
+ {
1437
+ "match": "kimi\u002dk3.1\u002dpro",
1438
+ "ref": "kimi\u002dk3.1\u002dpro"
1439
+ },
1440
+ {
1441
+ "match": "k3.1\u002dpro",
1442
+ "ref": "kimi\u002dk3.1\u002dpro"
1443
+ },
1444
+ {
1445
+ "match": "k3.1pro",
1446
+ "ref": "kimi\u002dk3.1\u002dpro"
820
1447
  }
821
1448
  ]
822
1449
  }