v1.6.0

What's Changed

• fix(api): strip region suffix in log model filter by @steebchen in #2787
• fix(gateway): warn-log expected upstream stream close by @steebchen in #2788
• fix: provider keys search, support notify, docs indent by @smakosh in #2790
• fix(models): correct MiMo cache read prices by @steebchen in #2792
• fix(gateway): keep reasoning on unpinned routing by @steebchen in #2791
• fix(gateway): synthesize Bedrock toolConfig from history by @steebchen in #2784
• fix(gateway): retry Azure missing-deployment 400 by @steebchen in #2796
• fix(gateway): reject TTS/video models in chat by @steebchen in #2793
• fix(gateway): use release date for models created field by @steebchen in #2797
• feat(admin): add client error and has-error log filters by @steebchen in #2798
• fix(actions): fall back when no region validation model by @steebchen in #2795
• fix(models): require releasedAt on every model by @steebchen in #2799
• chore(deps): bump actions/checkout from 6.0.3 to 7.0.0 by @dependabot[bot] in #2800
• chore(deps): bump pnpm/action-setup from 6.0.8 to 6.0.9 by @dependabot[bot] in #2801
• docs(blog): coding plans + chatgpt alternatives by @smakosh in #2794
• feat(models): add Sakana fugu-ultra model by @steebchen in #2802
• feat(model-selector): add region selection feature to model entries by @amelitoalcuitas in #2751
• fix(code): remove SoulForge cost/usage claims by @smakosh in #2804
• feat(ui): refresh comparisons + fix seo issues by @smakosh in #2805
• fix(security): patch js-yaml DoS via gray-matter by @steebchen in #2807
• feat(admin): model view toggle on cost breakdown by @steebchen in #2803
• fix(gateway): recover tool_call_id for Responses API by @steebchen in #2749
• feat(admin): add 10k/100k/1m max-log tiers by @steebchen in #2810
• fix(zai): map content filter + drop temperature by @steebchen in #2809
• fix(gateway): don't error on empty length-limited response by @steebchen in #2811
• feat(ui): add monetization & showcase templates by @smakosh in #2815
• docs: document Chat Playground port 3003 by @steebchen in #2817
• feat(logs): track requested vs used service tier by @steebchen in #2816
• fix(gateway): drop message name in Responses API input by @steebchen in #2820
• fix(bedrock): handle deprecated params for Opus 4.8 by @steebchen in #2813
• feat(gateway): add mistral-ocr-latest via /v1/ocr by @steebchen in #2818
• feat(gateway): gate service tiers to upstream provider keys by @steebchen in #2821
• fix(gateway): pricing from active provider in /v1/models by @steebchen in #2822
• fix(gateway): silence MALFORMED_RESPONSE finish reason by @steebchen in #2825
• fix(gateway): serve and expose Vertex service tiers by @steebchen in #2819
• feat: devpass growth loops + OCR OG image fix by @smakosh in #2827
• feat(models): add Granite stealth provider by @steebchen in #2824
• feat(gateway): gate routes by model output capability by @steebchen in #2828
• fix(api): zero-fill missing days in /activity range by @steebchen in #2829
• fix(gateway): guard recursive $ref schema expansion by @steebchen in #2830
• feat(chat): implement OCR message handling by @amelitoalcuitas in #2831
• fix(billing): prorate dev plan upgrades by @steebchen in #2566
• fix(devpass): match default billing org by email by @steebchen in #2837
• feat(code): show model family icon on coding cards by @steebchen in #2834
• feat(code): show DevPass invoice history by @steebchen in #2836
• fix(gateway): handle body-read socket close by @steebchen in #2826
• feat: org analytics + devpass profile updates by @smakosh in #2833
• fix(devpass): email invoice on plan upgrade by @steebchen in #2840
• feat(timeline): per-year hub redesign + SEO/AEO by @smakosh in #2843
• feat(chat): improve chat plan pricing & funnel by @smakosh in #2845
• chore(models): deprecate GLM-5.1 on Together AI by @steebchen in #2844
• feat(chat): add admin link to escalation email by @steebchen in #2847
• fix(gateway): forward image inputs to AWS Bedrock by @steebchen in #2846
• feat(scripts): weekly+monthly traffic Discord report by @smakosh in #2850
• fix(moonshot): deactivate retired kimi-k2-thinking line by @steebchen in #2849
• feat(models): add Tundra stealth provider by @steebchen in #2823
• fix(gateway): detect ai studio streaming served tier by @steebchen in #2851
• fix(admin): source provider/model/mapping history from mapping tables by @steebchen in #2611
• feat(db): tag SQL queries with app for Query Insights by @steebchen in #2655
• feat(payments-sdk): add opt-in preview feature flag by @steebchen in #2737
• fix(gateway): honor cancellation during body read by @steebchen in #2838
• test: migrate vitest poolOptions + guard TEST_MODELS by @steebchen in #2853

Full Changelog: v1.5.0...v1.6.0